# AI Gateway Configuration

Choose the request path before configuring authentication: Cloudflare REST inference, a Workers binding, and provider-native gateway endpoints have different requirements. Gateway access does not by itself define which upstream credentials or billing source a request uses.

| Task | Current documentation |
|------|-----------------------|
| Send a first request and locate the account/gateway identifiers | [Getting started](https://developers.cloudflare.com/ai-gateway/get-started/) |
| Create, update, or delete a gateway; use the default gateway | [Manage gateways](https://developers.cloudflare.com/ai-gateway/configuration/manage-gateway/) |
| Manage gateways programmatically | [Gateway management API](https://developers.cloudflare.com/api/resources/ai_gateway/methods/list/) |
| Protect gateway access and choose token permissions for the endpoint | [Authenticated Gateway](https://developers.cloudflare.com/ai-gateway/configuration/authentication/) and [REST API authentication](https://developers.cloudflare.com/ai-gateway/usage/rest-api/#authentication) |
| Configure Wrangler and an AI binding | [Workers AI binding setup](https://developers.cloudflare.com/ai-gateway/integrations/aig-workers-ai-binding/) and [binding methods](https://developers.cloudflare.com/ai-gateway/usage/worker-binding-methods/) |
| Store provider keys, select aliases, or diagnose missing credentials | [Bring Your Own Keys](https://developers.cloudflare.com/ai-gateway/configuration/bring-your-own-keys/) |
| Use Cloudflare billing or determine which credentials take precedence | [Unified Billing](https://developers.cloudflare.com/ai-gateway/features/unified-billing/) |
| Supply provider credentials with each request | [Provider guides](https://developers.cloudflare.com/ai-gateway/usage/providers/) |

Keep credentials out of source code. Use the chosen endpoint's documentation for headers and permissions instead of reusing an authentication recipe from another endpoint. Check credential precedence before changing an existing BYOK or billing setup.

For SDK selection, see [SDK integration](./sdk-integration.md); for policy settings, see [features](./features.md).

# AI Gateway Dynamic Routing

Use a dynamic route when model selection, traffic splitting, quotas, or fallbacks should be controlled in the gateway. For a simple retry or fallback sequence, check the request-handling and fallback guides before introducing a routing flow.

| Task | Current documentation |
|------|-----------------------|
| Design conditional routes, A/B splits, quotas, model fallbacks, and version rollbacks | [Dynamic routing](https://developers.cloudflare.com/ai-gateway/features/dynamic-routing/) |
| Invoke a route from an SDK, HTTP request, or Worker; inspect route response metadata | [Using a dynamic route](https://developers.cloudflare.com/ai-gateway/features/dynamic-routing/usage/) |
| Define route elements and connections programmatically | [JSON configuration](https://developers.cloudflare.com/ai-gateway/features/dynamic-routing/json-configuration/) |
| Provide metadata used by routing conditions | [Custom metadata](https://developers.cloudflare.com/ai-gateway/observability/custom-metadata/) |
| Configure model/provider fallbacks | [Fallbacks](https://developers.cloudflare.com/ai-gateway/configuration/fallbacks/) |
| Configure retries, backoff, and timeouts | [Request handling](https://developers.cloudflare.com/ai-gateway/configuration/request-handling/) |
| Inspect request outcomes, costs, and errors | [Analytics](https://developers.cloudflare.com/ai-gateway/observability/analytics/) and [logging](https://developers.cloudflare.com/ai-gateway/observability/logging/) |

Check the usage guide's authentication and stored-key prerequisites. Dynamic routes still use the [Unified API compatibility endpoint](https://developers.cloudflare.com/ai-gateway/usage/chat-completion/); its single-model deprecation does not make the REST inference endpoint a replacement for route invocation.

# AI Gateway Features

Fetch the relevant guide before setting feature flags, headers, limits, or billing behavior.

| Task | Current documentation |
|------|-----------------------|
| Enable caching, set TTLs, bypass the cache, or choose a custom cache key | [Caching](https://developers.cloudflare.com/ai-gateway/features/caching/) |
| Control request volume with fixed or sliding limits | [Rate limiting](https://developers.cloudflare.com/ai-gateway/features/rate-limiting/) |
| Enforce cost budgets | [Spend limits](https://developers.cloudflare.com/ai-gateway/features/spend-limits/) |
| Evaluate and enforce content policies | [Guardrails setup](https://developers.cloudflare.com/ai-gateway/features/guardrails/set-up-guardrail/) and [usage considerations](https://developers.cloudflare.com/ai-gateway/features/guardrails/usage-considerations/) |
| Detect sensitive data in prompts and responses | [DLP setup](https://developers.cloudflare.com/ai-gateway/features/dlp/set-up-dlp/) |
| Choose provider keys or Cloudflare billing | [BYOK](https://developers.cloudflare.com/ai-gateway/configuration/bring-your-own-keys/) and [Unified Billing](https://developers.cloudflare.com/ai-gateway/features/unified-billing/) |
| Configure provider data retention for Unified Billing | [Zero Data Retention](https://developers.cloudflare.com/ai-gateway/features/unified-billing/#zero-data-retention-zdr) |
| Configure log collection, payload storage, and retention | [Logging](https://developers.cloudflare.com/ai-gateway/observability/logging/) |
| Export logs | [Workers Logpush](https://developers.cloudflare.com/ai-gateway/observability/logging/logpush/) |
| Attach request metadata for tracking and routing | [Custom metadata](https://developers.cloudflare.com/ai-gateway/observability/custom-metadata/) |
| Override model costs | [Custom costs](https://developers.cloudflare.com/ai-gateway/configuration/custom-costs/) |
| Check supported providers, quotas, or pricing | [Provider guides](https://developers.cloudflare.com/ai-gateway/usage/providers/), [limits](https://developers.cloudflare.com/ai-gateway/reference/limits/), and [pricing](https://developers.cloudflare.com/ai-gateway/reference/pricing/) |

Choose cache keys only for requests whose responses may safely be shared. Decide what prompt and response data may be stored before enabling logging; do not infer a provider's retention policy from gateway log settings.

For conditional policies and fallbacks, see [dynamic routing](./dynamic-routing.md).

# Cloudflare AI Gateway

Use AI Gateway to observe and control requests to AI providers through caching, rate limiting, logging, and routing.

Fetch the linked documentation before choosing endpoints, authentication headers, SDK options, model names, or limits. Keep implementation details in the current docs.

## Choose a task

| Task | Reference |
|------|-----------|
| Create a gateway or choose authentication and provider credentials | [Configuration](./configuration.md) |
| Integrate an SDK, direct HTTP, or a Worker binding | [SDK integration](./sdk-integration.md) |
| Configure caching, rate limits, security, billing, or logging | [Features](./features.md) |
| Add fallbacks, conditional routing, or traffic splits | [Dynamic routing](./dynamic-routing.md) |
| Diagnose failed requests, caching, or missing logs | [Troubleshooting](./troubleshooting.md) |

For new single-model calls, start with the [REST API](https://developers.cloudflare.com/ai-gateway/usage/rest-api/) or [Workers bindings](https://developers.cloudflare.com/ai-gateway/usage/worker-binding-methods/), depending on the runtime. Preserve provider-native integrations when their API shape is needed; use the corresponding [provider guide](https://developers.cloudflare.com/ai-gateway/usage/providers/).

The [legacy Unified API](https://developers.cloudflare.com/ai-gateway/usage/chat-completion/) is deprecated for single-model calls but remains required for dynamic routes. Check the task before changing an existing endpoint.

Gateway authentication and upstream provider credentials are separate concerns. Choose the endpoint first, then follow its authentication and billing requirements in [configuration](./configuration.md).

## Related references

- [Workers AI](../workers-ai/README.md) — model inference.
- [Agents SDK documentation](https://developers.cloudflare.com/agents/) — stateful agents.
- [Vectorize](../vectorize/README.md) — vector search.

# AI Gateway SDK Integration

Choose an integration that matches the application's runtime and required API shape. Fetch its guide before installing packages or writing requests; model identifiers, SDK options, and gateway headers belong in the docs.

| Integration task | Current documentation |
|------------------|-----------------------|
| New single-model HTTP calls or OpenAI-compatible clients | [REST API](https://developers.cloudflare.com/ai-gateway/usage/rest-api/) |
| Vercel AI SDK, provider adapters, and fallback providers | [Vercel AI SDK integration](https://developers.cloudflare.com/ai-gateway/integrations/vercel-ai-sdk/) |
| Preserve OpenAI-native endpoints with the OpenAI SDK, Python, or HTTP | [OpenAI provider guide](https://developers.cloudflare.com/ai-gateway/usage/providers/openai/) |
| Preserve Anthropic-native requests | [Anthropic provider guide](https://developers.cloudflare.com/ai-gateway/usage/providers/anthropic/) |
| Use another provider or a framework's configurable provider endpoint | [Provider guides](https://developers.cloudflare.com/ai-gateway/usage/providers/) — match the framework's expected API shape to the provider endpoint |
| Configure Workers AI and an AI binding | [Binding setup](https://developers.cloudflare.com/ai-gateway/integrations/aig-workers-ai-binding/) |
| Call Workers AI or third-party models from a Worker; use gateway methods | [Workers binding methods](https://developers.cloudflare.com/ai-gateway/usage/worker-binding-methods/) |
| Set request metadata, caching, or other gateway headers | [Header glossary](https://developers.cloudflare.com/ai-gateway/glossary/) |
| Invoke a dynamic route | [Dynamic route usage](https://developers.cloudflare.com/ai-gateway/features/dynamic-routing/usage/) |

The [legacy Unified API](https://developers.cloudflare.com/ai-gateway/usage/chat-completion/) is deprecated for single-model calls, but dynamic routes still require its compatibility endpoint. Do not migrate a dynamic route to the REST inference endpoint as though it were a single-model call.

Confirm [gateway authentication and provider credentials](./configuration.md) separately, including the selected path's BYOK and billing behavior.

# AI Gateway Troubleshooting

Identify the request path and whether the failure comes from gateway access, upstream provider authentication, or request policy before changing credentials or retry behavior. A status code alone does not establish the failing layer.

| Symptom or task | Current documentation |
|-----------------|-----------------------|
| Authentication or provider errors, timeouts, DLP failures, or unexpected cache behavior | [Troubleshooting](https://developers.cloudflare.com/ai-gateway/reference/troubleshooting/) |
| Gateway authentication failure | [Authenticated Gateway](https://developers.cloudflare.com/ai-gateway/configuration/authentication/) and [REST API authentication](https://developers.cloudflare.com/ai-gateway/usage/rest-api/#authentication) |
| Provider key or billing mismatch | [BYOK](https://developers.cloudflare.com/ai-gateway/configuration/bring-your-own-keys/), [Unified Billing](https://developers.cloudflare.com/ai-gateway/features/unified-billing/), and the [provider guide](https://developers.cloudflare.com/ai-gateway/usage/providers/) |
| Rate limits or repeated failures | [Rate limiting](https://developers.cloudflare.com/ai-gateway/features/rate-limiting/) and [request handling](https://developers.cloudflare.com/ai-gateway/configuration/request-handling/) |
| Unexpected cache hit or miss, including streaming behavior | [Caching](https://developers.cloudflare.com/ai-gateway/features/caching/) |
| Missing logs, collection overrides, or storage limits | [Logging](https://developers.cloudflare.com/ai-gateway/observability/logging/) and [limits](https://developers.cloudflare.com/ai-gateway/reference/limits/) |
| Inspect headers, request metadata, usage, or export logs | [Header glossary](https://developers.cloudflare.com/ai-gateway/glossary/), [custom metadata](https://developers.cloudflare.com/ai-gateway/observability/custom-metadata/), [analytics](https://developers.cloudflare.com/ai-gateway/observability/analytics/), and [Logpush](https://developers.cloudflare.com/ai-gateway/observability/logging/logpush/) |
| Dynamic route failure | [Dynamic route usage](https://developers.cloudflare.com/ai-gateway/features/dynamic-routing/usage/) |

Check the existing SDK and gateway retry settings together before adding another retry loop. Follow [SDK integration](./sdk-integration.md) when an endpoint or model format is suspect.

# AI Search API Routes

Choose documentation that matches the integration you are working on. Fetch the reference before writing binding configuration, request types, response parsing, or streaming code.

| Task | Documentation |
|------|---------------|
| New Worker integration: bindings, search, chat completions, and streaming | [Search Workers binding](https://developers.cloudflare.com/ai-search/api/search/workers-binding/) |
| Create, list, configure, and inspect instances | [Instances Workers binding](https://developers.cloudflare.com/ai-search/api/instances/workers-binding/) |
| Query over HTTP and configure request authentication | [Search REST API](https://developers.cloudflare.com/ai-search/api/search/rest-api/) |
| Maintain an existing `env.AI.autorag()` integration | [Legacy Workers binding](https://developers.cloudflare.com/ai-search/api/migration/workers-binding-legacy/) |
| Upgrade a legacy binding, including responses, streaming, and filters | [Workers binding migration](https://developers.cloudflare.com/ai-search/api/migration/workers-binding/) |

The legacy binding remains supported; use current bindings for new integrations. Keep legacy request and response handling together until deliberately migrating them.

# AI Search Configuration Routes

| Task | Documentation |
|------|---------------|
| Set up a Worker and local development | [Workers binding quick start](https://developers.cloudflare.com/ai-search/get-started/workers/) |
| Select built-in uploads, R2, or a website; check supported formats | [Data sources](https://developers.cloudflare.com/ai-search/configuration/data-source/) |
| Connect existing R2 content | [R2 data source](https://developers.cloudflare.com/ai-search/configuration/data-source/r2/) |
| Configure crawling and diagnose bot-protection requirements | [Website data source](https://developers.cloudflare.com/ai-search/configuration/data-source/website/) |
| Include or exclude files and URL paths | [Path filtering](https://developers.cloudflare.com/ai-search/configuration/indexing/path-filtering/) |
| Configure source syncs, trigger indexing, or pause and resume | [Syncing](https://developers.cloudflare.com/ai-search/configuration/indexing/syncing/) |
| Grant AI Search access to R2 for indexing | [Service API token](https://developers.cloudflare.com/ai-search/configuration/indexing/service-api-token/) |
| Organize instances by application, tenant, or environment | [Namespaces](https://developers.cloudflare.com/ai-search/concepts/namespaces/) |
| Configure models | [Models](https://developers.cloudflare.com/ai-search/configuration/models/) |
| Inspect instance configuration and indexing progress | [Instances Workers binding](https://developers.cloudflare.com/ai-search/api/instances/workers-binding/) |

Indexing credentials and request authentication serve different purposes. For authenticating search requests over HTTP, use the [Search REST API](https://developers.cloudflare.com/ai-search/api/search/rest-api/) documentation.

# AI Search Troubleshooting Routes

| Symptom or question | Documentation |
|---------------------|---------------|
| API request fails, including authentication or missing instances | [API error codes](https://developers.cloudflare.com/ai-search/troubleshooting/api-error-codes/) |
| Upload or sync succeeds but content fails during processing | [Indexing error codes](https://developers.cloudflare.com/ai-search/troubleshooting/indexing-error-codes/) |
| Content is missing or stale | [Syncing](https://developers.cloudflare.com/ai-search/configuration/indexing/syncing/) and [supported data sources and formats](https://developers.cloudflare.com/ai-search/configuration/data-source/) |
| Filters return unexpected documents or no matches | [Filtering](https://developers.cloudflare.com/ai-search/configuration/retrieval/filtering/) and [metadata attributes](https://developers.cloudflare.com/ai-search/configuration/indexing/metadata/) |
| Thresholds exclude results or responses need tuning | [Result controls](https://developers.cloudflare.com/ai-search/configuration/retrieval/result-controls/) |
| Binding types, response parsing, or streaming fail after an upgrade | [Workers binding migration](https://developers.cloudflare.com/ai-search/api/migration/workers-binding/) |
| Capacity, file-size, or billing questions | [Limits and pricing](https://developers.cloudflare.com/ai-search/platform/limits-pricing/) |

For legacy binding behavior, start with [API routes](api.md). Do not apply current filter syntax or response shapes to legacy calls without following the migration guide.

# AI Search Pattern Routes

Choose retrieval-only search when your application displays chunks or handles generation itself; choose chat completions when AI Search should also generate the answer. Read [Search Workers binding](https://developers.cloudflare.com/ai-search/api/search/workers-binding/) for both paths and streaming behavior.

| Task | Documentation |
|------|---------------|
| Isolate tenants using separate instances or a shared filtered instance | [Multitenancy](https://developers.cloudflare.com/ai-search/how-to/per-tenant-search/) |
| Define built-in or custom metadata | [Metadata attributes](https://developers.cloudflare.com/ai-search/configuration/indexing/metadata/) |
| Filter by metadata, combine conditions, or match a folder and subfolders | [Filtering](https://developers.cloudflare.com/ai-search/configuration/retrieval/filtering/) |
| Tune result count and relevance thresholds | [Result controls](https://developers.cloudflare.com/ai-search/configuration/retrieval/result-controls/) |
| Resolve follow-up queries using conversation context | [Query rewriting](https://developers.cloudflare.com/ai-search/configuration/retrieval/query-rewriting/) |
| Improve result ordering with a second model | [Reranking](https://developers.cloudflare.com/ai-search/configuration/retrieval/reranking/) |
| Customize generation and query-rewriting instructions | [System prompt](https://developers.cloudflare.com/ai-search/configuration/retrieval/system-prompt/) |

For tenant isolation, read the full multitenancy guide before choosing an approach. A lower-bound folder comparison alone does not establish a tenant boundary; use the documented filtering semantics.

# Cloudflare AI Search

Use AI Search for managed content indexing and retrieval, with optional answer generation. Start with [How AI Search works](https://developers.cloudflare.com/ai-search/concepts/how-ai-search-works/).

## Choose the right product

- **Managed search or RAG over your content:** AI Search.
- **Custom embeddings and vector-index management:** [Vectorize](../vectorize/README.md).
- **Model inference without a managed retrieval pipeline:** [Workers AI](../workers-ai/README.md).

For freshness requirements, read [Syncing](https://developers.cloudflare.com/ai-search/configuration/indexing/syncing/) for your data source before choosing an architecture. Use the current [limits and pricing](https://developers.cloudflare.com/ai-search/platform/limits-pricing/) instead of assuming a fixed indexing interval or account limit.

## Find the right documentation

Read the linked page before implementing; these references route to the maintained documentation instead of copying API examples or configuration.

| Task | Start here |
|------|------------|
| Build a new Worker integration | [Workers binding quick start](https://developers.cloudflare.com/ai-search/get-started/workers/) |
| Choose an API or maintain an existing integration | [API routes](api.md) |
| Connect data, configure indexing, or manage environments | [Configuration routes](configuration.md) |
| Choose retrieval, generation, or tenant isolation patterns | [Pattern routes](patterns.md) |
| Diagnose indexing, authentication, filters, or limits | [Troubleshooting routes](gotchas.md) |

Existing `env.AI.autorag()` integrations can continue to work. Use the [migration guide](https://developers.cloudflare.com/ai-search/api/migration/workers-binding/) when upgrading; migration is not required just to maintain an existing integration.

# Analytics Engine API Reference

## Writing Data

### `writeDataPoint()`

Fire-and-forget (returns `void`, not Promise). Writes happen asynchronously.

```typescript
interface AnalyticsEngineDataPoint {
  blobs?: string[];      // Up to 20 strings (dimensions), 16KB each
  doubles?: number[];    // Up to 20 numbers (metrics)
  indexes?: string[];    // 1 indexed string for high-cardinality filtering
}

env.ANALYTICS.writeDataPoint({
  blobs: ["/api/users", "GET", "200"],
  doubles: [145.2, 1],  // latency_ms, count
  indexes: ["customer_abc123"]
});
```

**Behaviors:** No await needed, no error thrown (check tail logs), auto-sampled at high volumes, auto-timestamped.

**Blob vs Index:** Blob for GROUP BY (<100k unique), Index for filter-only (millions unique).

### Full Example

```typescript
export default {
  async fetch(request: Request, env: Env): Promise<Response> {
    const start = Date.now();
    const url = new URL(request.url);
    try {
      const response = await handleRequest(request);
      env.ANALYTICS.writeDataPoint({
        blobs: [url.pathname, request.method, response.status.toString()],
        doubles: [Date.now() - start, 1],
        indexes: [request.headers.get("x-api-key") || "anonymous"]
      });
      return response;
    } catch (error) {
      env.ANALYTICS.writeDataPoint({
        blobs: [url.pathname, request.method, "500"],
        doubles: [Date.now() - start, 1, 0],
      });
      throw error;
    }
  }
};
```

## SQL API (External Only)

```bash
curl -X POST https://api.cloudflare.com/client/v4/accounts/{account_id}/analytics_engine/sql \
  -H "Authorization: Bearer $TOKEN" \
  -d "SELECT blob1 AS endpoint, COUNT(*) AS requests FROM dataset WHERE timestamp >= NOW() - INTERVAL '1' HOUR GROUP BY blob1"
```

### Column References

```sql
-- blob1..blob20, double1..double20, index1, timestamp
SELECT blob1 AS endpoint, SUM(double1) AS latency, COUNT(*) AS requests
FROM my_dataset
WHERE index1 = 'customer_123' AND timestamp >= NOW() - INTERVAL '7' DAY
GROUP BY blob1
HAVING COUNT(*) > 100
ORDER BY requests DESC LIMIT 100
```

**Aggregations:** `SUM()`, `AVG()`, `COUNT()`, `MIN()`, `MAX()`, `quantile(0.95)()`

**Time ranges:** `NOW() - INTERVAL '1' HOUR`, `BETWEEN '2026-01-01' AND '2026-01-31'`

### Query Examples

```sql
-- Top endpoints
SELECT blob1, COUNT(*) AS requests, AVG(double1) AS avg_latency
FROM api_requests WHERE timestamp >= NOW() - INTERVAL '24' HOUR
GROUP BY blob1 ORDER BY requests DESC LIMIT 20

-- Error rate
SELECT blob1, COUNT(*) AS total,
  SUM(if(blob3 LIKE '5%', 1, 0)) AS errors
FROM api_requests WHERE timestamp >= NOW() - INTERVAL '1' HOUR
GROUP BY blob1 HAVING total > 50

-- P95 latency
SELECT blob1, quantile(0.95)(double1) AS p95
FROM api_requests GROUP BY blob1
```

## Response Format

```json
{"data": [{"endpoint": "/api/users", "requests": 1523}], "rows": 2}
```

## Limits

| Resource | Limit |
|----------|-------|
| Blobs/Doubles per point | 20 each |
| Indexes per point | 1 |
| Blob/Index size | 16KB |
| Data retention | 90 days |
| Query timeout | 30s |

**Critical:** High write volumes (>1M/min) trigger automatic sampling.

# Analytics Engine Configuration

## Setup

1. Add binding to `wrangler.jsonc`
2. Deploy Worker
3. Dataset created automatically on first write
4. Query via SQL API

## wrangler.jsonc

```jsonc
{
  "name": "my-worker",
  "analytics_engine_datasets": [
    { "binding": "ANALYTICS", "dataset": "my_events" }
  ]
}
```

Multiple datasets for separate concerns:
```jsonc
{
  "analytics_engine_datasets": [
    { "binding": "API_ANALYTICS", "dataset": "api_requests" },
    { "binding": "USER_EVENTS", "dataset": "user_activity" }
  ]
}
```

## TypeScript

```typescript
interface Env {
  ANALYTICS: AnalyticsEngineDataset;
}

export default {
  async fetch(request: Request, env: Env) {
    // No await - returns void, fire-and-forget
    env.ANALYTICS.writeDataPoint({
      blobs: [pathname, method, status],      // String dimensions (max 20)
      doubles: [latency, 1],                   // Numeric metrics (max 20)
      indexes: [apiKey]                        // High-cardinality filter (max 1)
    });
    return response;
  }
};
```

## Data Point Limits

| Field | Limit | SQL Access |
|-------|-------|------------|
| blobs | 20 strings, 16KB each | `blob1`...`blob20` |
| doubles | 20 numbers | `double1`...`double20` |
| indexes | 1 string, 16KB | `index1` |

## Write Behavior

| Scenario | Behavior |
|----------|----------|
| <1M writes/min | All accepted |
| >1M writes/min | Automatic sampling |
| Invalid data | Silent failure (check tail logs) |

**Mitigate sampling:** Pre-aggregate, use multiple datasets, write only critical metrics.

## Query Limits

| Resource | Limit |
|----------|-------|
| Query timeout | 30 seconds |
| Data retention | 90 days (default) |
| Result size | ~10MB |

## Cost

**Free tier:** 10M writes/month, 1M reads/month

**Paid:** $0.05 per 1M writes, $1.00 per 1M reads

## Environment-Specific

```jsonc
{
  "analytics_engine_datasets": [
    { "binding": "ANALYTICS", "dataset": "prod_events" }
  ],
  "env": {
    "staging": {
      "analytics_engine_datasets": [
        { "binding": "ANALYTICS", "dataset": "staging_events" }
      ]
    }
  }
}
```

## Monitoring

```bash
npx wrangler tail  # Check for sampling/write errors
```

```sql
-- Check write activity
SELECT DATE_TRUNC('hour', timestamp) AS hour, COUNT(*) AS writes
FROM my_dataset
WHERE timestamp >= NOW() - INTERVAL '24' HOUR
GROUP BY hour
```

# Analytics Engine Gotchas

## Critical Issues

### Sampling at High Volumes

**Problem:** Queries return fewer points than written at >1M writes/min.

**Solution:**
```typescript
// Pre-aggregate before writing
let buffer = { count: 0, total: 0 };
buffer.count++; buffer.total += value;

// Write once per second instead of per request
if (Date.now() % 1000 === 0) {
  env.ANALYTICS.writeDataPoint({ doubles: [buffer.count, buffer.total] });
}
```

**Detection:** `npx wrangler tail` → look for "sampling enabled"

### writeDataPoint Returns void

```typescript
// ❌ Pointless await
await env.ANALYTICS.writeDataPoint({...});

// ✅ Fire-and-forget
env.ANALYTICS.writeDataPoint({...});
```

Writes can fail silently. Check tail logs.

### Index vs Blob

| Cardinality | Use | Example |
|-------------|-----|---------|
| Millions | **Index** | user_id, api_key |
| Hundreds | **Blob** | endpoint, status_code, country |

```typescript
// ✅ Correct
{ blobs: [method, path, status], indexes: [userId] }
```

### Can't Query from Workers

Query API requires HTTP auth. Use external service or cache in KV/D1.

### No Custom Timestamps

Auto-generated at write time. Store original in blob if needed.

## Common Errors

| Error | Fix |
|-------|-----|
| Binding not found | Check wrangler.jsonc, redeploy |
| No data in query | Wait 30s; check dataset name; check time range |
| Query timeout | Add time filter; use index for filtering |

## Limits

| Resource | Limit |
|----------|-------|
| Blobs per point | 20 |
| Doubles per point | 20 |
| Indexes per point | 1 |
| Blob/Index size | 16KB |
| Write rate (no sampling) | ~1M/min |
| Retention | 90 days |
| Query timeout | 30s |

## Best Practices

✅ Pre-aggregate at high volumes  
✅ Use index for high-cardinality (millions)  
✅ Always include time filter in queries  
✅ Design schema before coding  

❌ Don't await writeDataPoint  
❌ Don't use index for low-cardinality  
❌ Don't query without time range  
❌ Don't assume all writes succeed

# Analytics Engine Patterns

## Use Cases

| Use Case | Key Metrics | Index On |
|----------|-------------|----------|
| API Metering | requests, bytes, compute_units | api_key |
| Feature Usage | feature, action, duration | user_id |
| Error Tracking | error_type, endpoint, count | customer_id |
| Performance | latency_ms, cache_status | endpoint |
| A/B Testing | variant, conversions | user_id |

## API Metering (Billing)

```typescript
env.ANALYTICS.writeDataPoint({
  blobs: [pathname, method, status, tier],
  doubles: [1, computeUnits, bytes, latencyMs],
  indexes: [apiKey]
});

// Query: Monthly usage by customer
// SELECT index1 AS api_key, SUM(double2) AS compute_units
// FROM usage WHERE timestamp >= DATE_TRUNC('month', NOW()) GROUP BY index1
```

## Error Tracking

```typescript
env.ANALYTICS.writeDataPoint({
  blobs: [endpoint, method, errorName, errorMessage.slice(0, 1000)],
  doubles: [1, timeToErrorMs],
  indexes: [customerId]
});
```

## Performance Monitoring

```typescript
env.ANALYTICS.writeDataPoint({
  blobs: [pathname, method, cacheStatus, status],
  doubles: [latencyMs, 1],
  indexes: [userId]
});

// Query: P95 latency by endpoint
// SELECT blob1, quantile(0.95)(double1) AS p95_ms FROM perf GROUP BY blob1
```

## Anti-Patterns

| ❌ Wrong | ✅ Correct |
|----------|-----------|
| `await writeDataPoint()` | `writeDataPoint()` (fire-and-forget) |
| `indexes: [method]` (low cardinality) | `blobs: [method]`, `indexes: [userId]` |
| `blobs: [JSON.stringify(obj)]` | Store ID in blob, full object in D1/KV |
| Write every request at 10M/min | Pre-aggregate per second |
| Query from Worker | Query from external service/API |

## Best Practices

1. **Design schema upfront** - Document blob/double/index assignments
2. **Always include count metric** - `doubles: [latency, 1]` for AVG calculations
3. **Use enums for blobs** - Consistent values like `Status.SUCCESS`
4. **Handle sampling** - Use ratios (avg_latency = SUM(latency)/SUM(count))
5. **Test queries early** - Validate schema before heavy writes

## Schema Template

```typescript
/**
 * Dataset: my_metrics
 * 
 * Blobs:
 *   blob1: endpoint, blob2: method, blob3: status
 * 
 * Doubles:
 *   double1: latency_ms, double2: count (always 1)
 * 
 * Indexes:
 *   index1: customer_id (high cardinality)
 */
```

# Cloudflare Workers Analytics Engine Reference

Expert guidance for implementing unlimited-cardinality analytics at scale using Cloudflare Workers Analytics Engine.

## What is Analytics Engine?

Time-series analytics database designed for high-cardinality data (millions of unique dimensions). Write data points from Workers, query via SQL API. Use for:
- Custom user-facing analytics dashboards
- Usage-based billing & metering
- Per-customer/per-feature monitoring
- High-frequency instrumentation without performance impact

**Key Capability:** Track metrics with unlimited unique values (e.g., millions of user IDs, API keys) without performance degradation.

## Core Concepts

| Concept | Description | Example |
|---------|-------------|---------|
| **Dataset** | Logical table for related metrics | `api_requests`, `user_events` |
| **Data Point** | Single measurement with timestamp | One API request's metrics |
| **Blobs** | String dimensions (max 20) | endpoint, method, status, user_id |
| **Doubles** | Numeric values (max 20) | latency_ms, request_count, bytes |
| **Indexes** | Filtered blobs for efficient queries | customer_id, api_key |

## Reading Order

| Task | Start Here | Then Read |
|------|------------|-----------|
| **First-time setup** | [configuration.md](configuration.md) → [api.md](api.md) → [patterns.md](patterns.md) | |
| **Writing data** | [api.md](api.md) → [gotchas.md](gotchas.md) (sampling) | |
| **Querying data** | [api.md](api.md) (SQL API) → [patterns.md](patterns.md) (examples) | |
| **Debugging** | [gotchas.md](gotchas.md) → [api.md](api.md) (limits) | |
| **Optimization** | [patterns.md](patterns.md) (anti-patterns) → [gotchas.md](gotchas.md) | |

## When to Use Analytics Engine

```
Need to track metrics? → Yes
  ↓
Millions of unique dimension values? → Yes
    ↓
  Need real-time queries? → Yes
      ↓
    Use Analytics Engine ✓

Alternative scenarios:
- Low cardinality (<10k unique values) → Workers Analytics (free tier)
- Complex joins/relations → D1 Database
- Logs/debugging → Tail Workers (logpush)
- External tools → Send to external analytics (Datadog, etc.)
```

## Quick Start

1. Add binding to `wrangler.jsonc`:
```jsonc
{
  "analytics_engine_datasets": [
    { "binding": "ANALYTICS", "dataset": "my_events" }
  ]
}
```

2. Write data points (fire-and-forget, no await):
```typescript
env.ANALYTICS.writeDataPoint({
  blobs: ["/api/users", "GET", "200"],
  doubles: [145.2, 1],  // latency_ms, count
  indexes: [customerId]
});
```

3. Query via SQL API (HTTP):
```sql
SELECT blob1, SUM(double2) AS total_requests
FROM my_events
WHERE index1 = 'customer_123'
  AND timestamp >= NOW() - INTERVAL '7' DAY
GROUP BY blob1
ORDER BY total_requests DESC
```

## In This Reference

- **[configuration.md](configuration.md)** - Setup, bindings, TypeScript types, limits
- **[api.md](api.md)** - `writeDataPoint()`, SQL API, query syntax
- **[patterns.md](patterns.md)** - Use cases, examples, anti-patterns
- **[gotchas.md](gotchas.md)** - Sampling, index selection, troubleshooting

## See Also

- [Cloudflare Analytics Engine Docs](https://developers.cloudflare.com/analytics/analytics-engine/)
- [GraphQL Analytics API Reference](../graphql-api/) - Query built-in Cloudflare analytics (HTTP, Workers, DNS, Firewall, etc.)
- [Observability Reference](../observability/) - Workers Logs, Traces, and real-time debugging

# API Reference

## Client Initialization

### TypeScript

```typescript
import Cloudflare from 'cloudflare';

const client = new Cloudflare({
  apiToken: process.env.CLOUDFLARE_API_TOKEN,
});
```

### Python

```python
from cloudflare import Cloudflare

client = Cloudflare(api_token=os.environ.get("CLOUDFLARE_API_TOKEN"))

# For async:
from cloudflare import AsyncCloudflare
client = AsyncCloudflare(api_token=os.environ["CLOUDFLARE_API_TOKEN"])
```

### Go

```go
import (
    "github.com/cloudflare/cloudflare-go/v4"
    "github.com/cloudflare/cloudflare-go/v4/option"
)

client := cloudflare.NewClient(
    option.WithAPIToken(os.Getenv("CLOUDFLARE_API_TOKEN")),
)
```

## Authentication

### API Token (Recommended)

**Create token**: Dashboard → My Profile → API Tokens → Create Token

```bash
export CLOUDFLARE_API_TOKEN='your-token-here'

curl "https://api.cloudflare.com/client/v4/zones" \
  --header "Authorization: Bearer $CLOUDFLARE_API_TOKEN"
```

**Token scopes**: Always use minimal permissions (zone-specific, time-limited).

### API Key (Legacy)

```bash
curl "https://api.cloudflare.com/client/v4/zones" \
  --header "X-Auth-Email: user@example.com" \
  --header "X-Auth-Key: $CLOUDFLARE_API_KEY"
```

**Not recommended:** Full account access, cannot scope permissions.

## Auto-Pagination

All SDKs support automatic pagination for list operations.

```typescript
// TypeScript: for await...of
for await (const zone of client.zones.list()) {
  console.log(zone.id);
}
```

```python
# Python: iterator protocol
for zone in client.zones.list():
    print(zone.id)
```

```go
// Go: ListAutoPaging
iter := client.Zones.ListAutoPaging(ctx, cloudflare.ZoneListParams{})
for iter.Next() {
    zone := iter.Current()
    fmt.Println(zone.ID)
}
```

## Error Handling

```typescript
try {
  const zone = await client.zones.get({ zone_id: 'xxx' });
} catch (err) {
  if (err instanceof Cloudflare.NotFoundError) {
    // 404
  } else if (err instanceof Cloudflare.RateLimitError) {
    // 429 - SDK auto-retries with backoff
  } else if (err instanceof Cloudflare.APIError) {
    console.log(err.status, err.message);
  }
}
```

**Common Error Types:**
- `AuthenticationError` (401) - Invalid token
- `PermissionDeniedError` (403) - Insufficient scope
- `NotFoundError` (404) - Resource not found
- `RateLimitError` (429) - Rate limit exceeded
- `InternalServerError` (≥500) - Cloudflare error

## Zone Management

```typescript
// List zones
const zones = await client.zones.list({
  account: { id: 'account-id' },
  status: 'active',
});

// Create zone
const zone = await client.zones.create({
  account: { id: 'account-id' },
  name: 'example.com',
  type: 'full', // or 'partial'
});

// Update zone
await client.zones.edit('zone-id', {
  paused: false,
});

// Delete zone
await client.zones.delete('zone-id');
```

```go
// Go: requires cloudflare.F() wrapper
zone, err := client.Zones.New(ctx, cloudflare.ZoneNewParams{
    Account: cloudflare.F(cloudflare.ZoneNewParamsAccount{
        ID: cloudflare.F("account-id"),
    }),
    Name: cloudflare.F("example.com"),
    Type: cloudflare.F(cloudflare.ZoneNewParamsTypeFull),
})
```

## DNS Management

```typescript
// Create DNS record
await client.dns.records.create({
  zone_id: 'zone-id',
  type: 'A',
  name: 'subdomain.example.com',
  content: '192.0.2.1',
  ttl: 1, // auto
  proxied: true, // Orange cloud
});

// List DNS records (with auto-pagination)
for await (const record of client.dns.records.list({
  zone_id: 'zone-id',
  type: 'A',
})) {
  console.log(record.name, record.content);
}

// Update DNS record
await client.dns.records.update({
  zone_id: 'zone-id',
  dns_record_id: 'record-id',
  type: 'A',
  name: 'subdomain.example.com',
  content: '203.0.113.1',
  proxied: true,
});

// Delete DNS record
await client.dns.records.delete({
  zone_id: 'zone-id',
  dns_record_id: 'record-id',
});
```

```python
# Python example
client.dns.records.create(
    zone_id="zone-id",
    type="A",
    name="subdomain.example.com",
    content="192.0.2.1",
    ttl=1,
    proxied=True,
)
```

## See Also

- [configuration.md](./configuration.md) - SDK configuration, environment variables
- [patterns.md](./patterns.md) - Real-world patterns and workflows
- [gotchas.md](./gotchas.md) - Rate limits, troubleshooting

# Configuration

## Environment Variables

### Set Variables

| Platform | Command |
|----------|---------|
| Linux/macOS | `export CLOUDFLARE_API_TOKEN='token'` |
| PowerShell | `$env:CLOUDFLARE_API_TOKEN = 'token'` |
| Windows CMD | `set CLOUDFLARE_API_TOKEN=token` |

**Security:** Never commit tokens. Use `.env` files (gitignored) or secret managers.

### .env File Pattern

```bash
# .env (add to .gitignore)
CLOUDFLARE_API_TOKEN=your-token-here
CLOUDFLARE_ACCOUNT_ID=your-account-id
```

```typescript
// TypeScript
import 'dotenv/config';

const client = new Cloudflare({
  apiToken: process.env.CLOUDFLARE_API_TOKEN,
});
```

```python
# Python
from dotenv import load_dotenv
load_dotenv()

client = Cloudflare(api_token=os.environ["CLOUDFLARE_API_TOKEN"])
```

## SDK Configuration

### TypeScript

```typescript
const client = new Cloudflare({
  apiToken: process.env.CLOUDFLARE_API_TOKEN,
  timeout: 120000,        // 2 min (default 60s), in milliseconds
  maxRetries: 5,          // default 2
  baseURL: 'https://...', // proxy (rare)
});

// Per-request overrides
await client.zones.get(
  { zone_id: 'zone-id' },
  { timeout: 5000, maxRetries: 0 }
);
```

### Python

```python
client = Cloudflare(
    api_token=os.environ["CLOUDFLARE_API_TOKEN"],
    timeout=120,         # seconds (default 60)
    max_retries=5,       # default 2
    base_url="https://...",  # proxy (rare)
)

# Per-request overrides
client.with_options(timeout=5, max_retries=0).zones.get(zone_id="zone-id")
```

### Go

```go
client := cloudflare.NewClient(
    option.WithAPIToken(os.Getenv("CLOUDFLARE_API_TOKEN")),
    option.WithMaxRetries(5),  // default 10 (higher than TS/Python)
    option.WithRequestTimeout(2 * time.Minute),  // default 60s
    option.WithBaseURL("https://..."),  // proxy (rare)
)

// Per-request overrides
client.Zones.Get(ctx, "zone-id", option.WithMaxRetries(0))
```

## Configuration Options

| Option | TypeScript | Python | Go | Default |
|--------|-----------|--------|-----|---------|
| Timeout | `timeout` (ms) | `timeout` (s) | `WithRequestTimeout` | 60s |
| Retries | `maxRetries` | `max_retries` | `WithMaxRetries` | 2 (Go: 10) |
| Base URL | `baseURL` | `base_url` | `WithBaseURL` | api.cloudflare.com |

**Note:** Go SDK has higher default retries (10) than TypeScript/Python (2).

## Timeout Configuration

**When to increase:**
- Large zone transfers
- Bulk DNS operations
- Worker script uploads

```typescript
const client = new Cloudflare({
  timeout: 300000, // 5 minutes
});
```

## Retry Configuration

**When to increase:** Rate-limit-heavy workflows, flaky network

**When to decrease:** Fast-fail requirements, user-facing requests

```typescript
// Increase retries for batch operations
const client = new Cloudflare({ maxRetries: 10 });

// Disable retries for fast-fail
const fastClient = new Cloudflare({ maxRetries: 0 });
```

## Wrangler CLI Integration

```bash
# Configure authentication
wrangler login
# Or
export CLOUDFLARE_API_TOKEN='token'

# Common commands that use API
wrangler deploy              # Uploads worker via API
wrangler kv:key put          # KV operations
wrangler r2 bucket create    # R2 operations
wrangler d1 execute          # D1 operations
wrangler pages deploy        # Pages operations

# Get API configuration
wrangler whoami              # Shows authenticated user
```

### wrangler.toml

```toml
name = "my-worker"
main = "src/index.ts"
compatibility_date = "2024-01-01"
account_id = "your-account-id"

# Can also use env vars:
# CLOUDFLARE_ACCOUNT_ID
# CLOUDFLARE_API_TOKEN
```

## See Also

- [api.md](./api.md) - Client initialization, authentication
- [gotchas.md](./gotchas.md) - Rate limits, timeout errors
- [Wrangler Reference](https://developers.cloudflare.com/workers/wrangler/) - CLI tool details

# Gotchas & Troubleshooting

## Rate Limits & 429 Errors

**Actual Limits:**
- **1200 requests / 5 minutes** per user/token (global)
- **200 requests / second** per IP address
- **GraphQL: 320 / 5 minutes** (cost-based)

**SDK Behavior:**
- Auto-retry with exponential backoff (default 2 retries, Go: 10)
- Respects `Retry-After` header
- Throws `RateLimitError` after exhausting retries

**Solution:**

```typescript
// Increase retries for rate-limit-heavy workflows
const client = new Cloudflare({ maxRetries: 5 });

// Add application-level throttling
import pLimit from 'p-limit';
const limit = pLimit(10); // Max 10 concurrent requests
```

## SDK-Specific Issues

### Go: Required Field Wrapper

**Problem:** Go SDK requires `cloudflare.F()` wrapper for optional fields.

```go
// ❌ WRONG - Won't compile or send field
client.Zones.New(ctx, cloudflare.ZoneNewParams{
    Name: "example.com",
})

// ✅ CORRECT
client.Zones.New(ctx, cloudflare.ZoneNewParams{
    Name: cloudflare.F("example.com"),
    Account: cloudflare.F(cloudflare.ZoneNewParamsAccount{
        ID: cloudflare.F("account-id"),
    }),
})
```

**Why:** Distinguishes between zero value, null, and omitted fields.

### Python: Async vs Sync Clients

**Problem:** Using sync client in async context or vice versa.

```python
# ❌ WRONG - Can't await sync client
from cloudflare import Cloudflare
client = Cloudflare()
await client.zones.list()  # TypeError

# ✅ CORRECT - Use AsyncCloudflare
from cloudflare import AsyncCloudflare
client = AsyncCloudflare()
await client.zones.list()
```

## Token Permission Errors (403)

**Problem:** API returns 403 Forbidden despite valid token.

**Cause:** Token lacks required permissions (scope).

**Scopes Required:**

| Operation | Required Scope |
|-----------|----------------|
| List zones | Zone:Read (zone-level or account-level) |
| Create zone | Zone:Edit (account-level) |
| Edit DNS | DNS:Edit (zone-level) |
| Deploy Worker | Workers Script:Edit (account-level) |
| Read KV | Workers KV Storage:Read |
| Write KV | Workers KV Storage:Edit |

**Solution:** Re-create token with correct permissions in Dashboard → My Profile → API Tokens.

## Pagination Truncation

**Problem:** Only getting first 20 results (default page size).

**Solution:** Use auto-pagination iterators.

```typescript
// ❌ WRONG - Only first page (20 items)
const page = await client.zones.list();

// ✅ CORRECT - All results
const zones = [];
for await (const zone of client.zones.list()) {
  zones.push(zone);
}
```

## Workers Subrequests

**Problem:** Rate limit hit faster than expected in Workers.

**Cause:** Workers subrequests count as separate API calls.

**Solution:** Use bindings instead of REST API in Workers (see ../bindings/).

```typescript
// ❌ WRONG - REST API in Workers (counts against rate limit)
const client = new Cloudflare({ apiToken: env.CLOUDFLARE_API_TOKEN });
const zones = await client.zones.list();

// ✅ CORRECT - Use bindings (no rate limit)
// Access via env.MY_BINDING
```

## Authentication Errors (401)

**Problem:** "Authentication failed" or "Invalid token"

**Causes:**
- Token expired
- Token deleted/revoked
- Token not set in environment
- Wrong token format

**Solution:**

```typescript
// Verify token is set
if (!process.env.CLOUDFLARE_API_TOKEN) {
  throw new Error('CLOUDFLARE_API_TOKEN not set');
}

// Test token
const user = await client.user.tokens.verify();
console.log('Token valid:', user.status);
```

## Timeout Errors

**Problem:** Request times out (default 60s).

**Cause:** Large operations (bulk DNS, zone transfers).

**Solution:** Increase timeout or split operations.

```typescript
// Increase timeout
const client = new Cloudflare({
  timeout: 300000, // 5 minutes
});

// Or split operations
const batchSize = 100;
for (let i = 0; i < records.length; i += batchSize) {
  const batch = records.slice(i, i + batchSize);
  await processBatch(batch);
}
```

## Zone Not Found (404)

**Problem:** Zone ID valid but returns 404.

**Causes:**
- Zone not in account associated with token
- Zone deleted
- Wrong zone ID format

**Solution:**

```typescript
// List all zones to find correct ID
for await (const zone of client.zones.list()) {
  console.log(zone.id, zone.name);
}
```

## Limits Reference

| Resource/Limit | Value | Notes |
|----------------|-------|-------|
| API rate limit | 1200/5min | Per user/token |
| IP rate limit | 200/sec | Per IP |
| GraphQL rate limit | 320/5min | Cost-based |
| Parallel requests (recommended) | < 10 | Avoid overwhelming API |
| Default page size | 20 | Use auto-pagination |
| Max page size | 50 | Some endpoints |

## Best Practices

**Security:**
- Never commit tokens
- Use minimal permissions
- Rotate tokens regularly
- Set token expiration

**Performance:**
- Batch operations
- Use pagination wisely
- Cache responses
- Handle rate limits

**Code Organization:**

```typescript
// Create reusable client instance
export const cfClient = new Cloudflare({
  apiToken: process.env.CLOUDFLARE_API_TOKEN,
  maxRetries: 5,
});

// Wrap common operations
export async function getZoneDetails(zoneId: string) {
  return await cfClient.zones.get({ zone_id: zoneId });
}
```

## See Also

- [api.md](./api.md) - Error types, authentication
- [configuration.md](./configuration.md) - Timeout/retry configuration
- [patterns.md](./patterns.md) - Error handling patterns

# Common Patterns

## List All with Auto-Pagination

**Problem:** API returns paginated results. Default page size is 20.

**Solution:** Use SDK auto-pagination to iterate all results.

```typescript
// TypeScript
for await (const zone of client.zones.list()) {
  console.log(zone.name);
}
```

```python
# Python
for zone in client.zones.list():
    print(zone.name)
```

```go
// Go
iter := client.Zones.ListAutoPaging(ctx, cloudflare.ZoneListParams{})
for iter.Next() {
    fmt.Println(iter.Current().Name)
}
```

## Error Handling with Retry

**Problem:** Rate limits (429) and transient errors need retry.

**Solution:** SDKs auto-retry with exponential backoff. Customize as needed.

```typescript
// Increase retries for rate-limit-heavy operations
const client = new Cloudflare({ maxRetries: 5 });

try {
  const zone = await client.zones.create({ /* ... */ });
} catch (err) {
  if (err instanceof Cloudflare.RateLimitError) {
    // Already retried 5 times with backoff
    const retryAfter = err.headers['retry-after'];
    console.log(`Rate limited. Retry after ${retryAfter}s`);
  }
}
```

## Batch Parallel Operations

**Problem:** Need to create multiple resources quickly.

**Solution:** Use `Promise.all()` for parallel requests (respect rate limits).

```typescript
// Create multiple DNS records in parallel
const records = ['www', 'api', 'cdn'].map(subdomain =>
  client.dns.records.create({
    zone_id: 'zone-id',
    type: 'A',
    name: `${subdomain}.example.com`,
    content: '192.0.2.1',
  })
);
await Promise.all(records);
```

**Controlled concurrency** (avoid rate limits):

```typescript
import pLimit from 'p-limit';
const limit = pLimit(10); // Max 10 concurrent

const subdomains = ['www', 'api', 'cdn', /* many more */];
const records = subdomains.map(subdomain =>
  limit(() => client.dns.records.create({
    zone_id: 'zone-id',
    type: 'A',
    name: `${subdomain}.example.com`,
    content: '192.0.2.1',
  }))
);
await Promise.all(records);
```

## Zone CRUD Workflow

```typescript
// Create
const zone = await client.zones.create({
  account: { id: 'account-id' },
  name: 'example.com',
  type: 'full',
});

// Read
const fetched = await client.zones.get({ zone_id: zone.id });

// Update
await client.zones.edit(zone.id, { paused: false });

// Delete
await client.zones.delete(zone.id);
```

## DNS Bulk Update

```typescript
// Fetch all A records
const records = [];
for await (const record of client.dns.records.list({
  zone_id: 'zone-id',
  type: 'A',
})) {
  records.push(record);
}

// Update all to new IP
await Promise.all(records.map(record =>
  client.dns.records.update({
    zone_id: 'zone-id',
    dns_record_id: record.id,
    type: 'A',
    name: record.name,
    content: '203.0.113.1', // New IP
    proxied: record.proxied,
    ttl: record.ttl,
  })
));
```

## Filter and Collect Results

```typescript
// Find all proxied A records
const proxiedRecords = [];
for await (const record of client.dns.records.list({
  zone_id: 'zone-id',
  type: 'A',
})) {
  if (record.proxied) {
    proxiedRecords.push(record);
  }
}
```

## Error Recovery Pattern

```typescript
async function createZoneWithRetry(name: string, maxAttempts = 3) {
  for (let attempt = 1; attempt <= maxAttempts; attempt++) {
    try {
      return await client.zones.create({
        account: { id: 'account-id' },
        name,
        type: 'full',
      });
    } catch (err) {
      if (err instanceof Cloudflare.RateLimitError && attempt < maxAttempts) {
        const retryAfter = parseInt(err.headers['retry-after'] || '5');
        console.log(`Rate limited, waiting ${retryAfter}s (retry ${attempt}/${maxAttempts})`);
        await new Promise(resolve => setTimeout(resolve, retryAfter * 1000));
      } else {
        throw err;
      }
    }
  }
}
```

## Conditional Update Pattern

```typescript
// Only update if zone is active
const zone = await client.zones.get({ zone_id: 'zone-id' });
if (zone.status === 'active') {
  await client.zones.edit(zone.id, { paused: false });
}
```

## Batch with Error Handling

```typescript
// Process multiple zones, continue on errors
const results = await Promise.allSettled(
  zoneIds.map(id => client.zones.get({ zone_id: id }))
);

results.forEach((result, i) => {
  if (result.status === 'fulfilled') {
    console.log(`Zone ${i}: ${result.value.name}`);
  } else {
    console.error(`Zone ${i} failed:`, result.reason.message);
  }
});
```

## See Also

- [api.md](./api.md) - SDK client initialization, basic operations
- [gotchas.md](./gotchas.md) - Rate limits, common errors
- [configuration.md](./configuration.md) - SDK configuration options

# Cloudflare API Integration

Guide for working with Cloudflare's REST API - authentication, SDK usage, common patterns, and troubleshooting.

## Quick Decision Tree

```
How are you calling the Cloudflare API?
├─ From Workers runtime → Use bindings, not REST API (see ../bindings/)
├─ Server-side (Node/Python/Go) → Official SDK (see api.md)
├─ CLI/scripts → Wrangler or curl (see configuration.md)
├─ Infrastructure-as-code → See ../pulumi/ or ../terraform/
└─ One-off requests → curl examples (see api.md)
```

## SDK Selection

| Language | Package | Best For | Default Retries |
|----------|---------|----------|-----------------|
| TypeScript | `cloudflare` | Node.js, Bun, Next.js, Workers | 2 |
| Python | `cloudflare` | FastAPI, Django, scripts | 2 |
| Go | `cloudflare-go/v4` | CLI tools, microservices | 10 |

All SDKs are Stainless-generated from OpenAPI spec (consistent APIs).

## Authentication Methods

| Method | Security | Use Case | Scope |
|--------|----------|----------|-------|
| **API Token** ✓ | Scoped, rotatable | Production | Per-zone or account |
| API Key + Email | Full account access | Legacy only | Everything |
| User Service Key | Limited | Origin CA certs only | Origin CA |

**Always use API tokens** for new projects.

## Rate Limits

| Limit | Value |
|-------|-------|
| Per user/token | 1200 requests / 5 minutes |
| Per IP | 200 requests / second |
| GraphQL | 320 / 5 minutes (cost-based) |

## Reading Order

| Task | Files to Read |
|------|---------------|
| Initialize SDK client | api.md |
| Configure auth/timeout/retry | configuration.md |
| Find usage patterns | patterns.md |
| Debug errors/rate limits | gotchas.md |
| Product-specific APIs | [Workers docs](https://developers.cloudflare.com/workers/), ../r2/, ../kv/, etc. |

## In This Reference

- **[api.md](api.md)** - SDK client initialization, pagination, error handling, examples
- **[configuration.md](configuration.md)** - Environment variables, SDK config, Wrangler setup
- **[patterns.md](patterns.md)** - Real-world patterns, batch operations, workflows
- **[gotchas.md](gotchas.md)** - Rate limits, SDK-specific issues, troubleshooting

## See Also

- [Cloudflare API Docs](https://developers.cloudflare.com/api/)
- [Bindings Reference](../bindings/) - Workers runtime bindings (preferred over REST API)
- [Wrangler Reference](https://developers.cloudflare.com/workers/wrangler/) - CLI tool for Cloudflare development
- [GraphQL Analytics API Reference](../graphql-api/) - Analytics data via GraphQL (separate endpoint from REST API)

# API Reference

Base: `/zones/{zone_id}/api_gateway`

## Endpoints

```bash
GET /operations                    # List
GET /operations/{op_id}            # Get single
POST /operations/item              # Create: {endpoint,host,method}
POST /operations                   # Bulk: {operations:[{endpoint,host,method}]}
DELETE /operations/{op_id}         # Delete
DELETE /operations                 # Bulk delete: {operation_ids:[...]}
```

## Discovery

```bash
GET /discovery/operations                    # List discovered
PATCH /discovery/operations/{op_id}          # Update: {state:"saved"|"ignored"}
PATCH /discovery/operations                  # Bulk: {operation_ids:{id:{state}}}
GET /discovery                               # OpenAPI export
```

## Config

```bash
GET /configuration        # Get session ID config
PUT /configuration        # Update: {auth_id_characteristics:[{name,type:"header"|"cookie"}]}
```

## Token Validation

```bash
GET /token_validation                  # List
POST /token_validation                 # Create: {name,location:{header:"..."},jwks:"..."}
POST /jwt_validation_rules             # Rule: {name,hostname,token_validation_id,action:"block"}
```

## Workers Integration

### Access JWT Claims
```js
export default {
  async fetch(req, env) {
    // Access validated JWT payload
    const jwt = req.cf?.jwt?.payload?.[env.JWT_CONFIG_ID]?.[0];
    if (jwt) {
      const userId = jwt.sub;
      const role = jwt.role;
    }
  }
}
```

### Access mTLS Info
```js
export default {
  async fetch(req, env) {
    const tls = req.cf?.tlsClientAuth;
    if (tls?.certVerified === 'SUCCESS') {
      const fingerprint = tls.certFingerprintSHA256;
      // Authenticated client
    }
  }
}
```

### Dynamic JWKS Update
```js
export default {
  async scheduled(event, env) {
    const jwks = await (await fetch('https://auth.example.com/.well-known/jwks.json')).json();
    await fetch(`https://api.cloudflare.com/client/v4/zones/${env.ZONE_ID}/api_gateway/token_validation/${env.CONFIG_ID}`, {
      method: 'PATCH',
      headers: {'Authorization': `Bearer ${env.CF_API_TOKEN}`, 'Content-Type': 'application/json'},
      body: JSON.stringify({jwks: JSON.stringify(jwks)})
    });
  }
}
```

## Firewall Fields

### Core Fields
```js
cf.api_gateway.auth_id_present           // Session ID present
cf.api_gateway.request_violates_schema   // Schema violation
cf.api_gateway.fallthrough_triggered     // No endpoint match
cf.tls_client_auth.cert_verified         // mTLS cert valid
cf.tls_client_auth.cert_fingerprint_sha256
```

### JWT Validation (2026)
```js
// Modern validation syntax
is_jwt_valid(http.request.jwt.payload["{config_id}"][0])

// Legacy (still supported)
cf.api_gateway.jwt_claims_valid

// Extract claims
lookup_json_string(http.request.jwt.payload["{config_id}"][0], "claim_name")
```

### Risk Labels (2026)
```js
// BOLA detection
cf.api_gateway.cf-risk-bola-enumeration  // Sequential resource access detected
cf.api_gateway.cf-risk-bola-pollution    // Parameter pollution detected

// Authentication posture
cf.api_gateway.cf-risk-missing-auth      // Endpoint lacks authentication
cf.api_gateway.cf-risk-mixed-auth        // Inconsistent auth patterns
```

## BOLA Detection

```bash
GET /user_schemas/{schema_id}/bola             # Get BOLA config
PATCH /user_schemas/{schema_id}/bola           # Update: {enabled:true}
```

## Auth Posture

```bash
GET /discovery/authentication_posture          # List unprotected endpoints
```

## GraphQL Protection

```bash
GET /settings/graphql_protection               # Get limits
PUT /settings/graphql_protection               # Set: {max_depth,max_size}
```

## See Also

- [configuration.md](configuration.md) - Setup guides for all features
- [patterns.md](patterns.md) - Firewall rules and common patterns
- [API Gateway API Docs](https://developers.cloudflare.com/api/resources/api_gateway/)

# Configuration

## Schema Validation 2.0 Setup

> ⚠️ **Classic Schema Validation deprecated.** Use Schema Validation 2.0.

**Upload schema (Dashboard):**
```
Security > API Shield > Schema Validation > Add validation
- Upload .yml/.yaml/.json (OpenAPI v3.0)
- Endpoints auto-added to Endpoint Management
- Action: Log | Block | None
- Body inspection: JSON payloads
```

**Change validation action:**
```
Security > API Shield > Settings > Schema Validation
Per-endpoint: Filter → ellipses → Change action
Default action: Set global mitigation action
```

**Migration from Classic:**
```
1. Export existing schema (if available)
2. Delete all Classic schema validation rules
3. Wait 5 min for cache clear
4. Re-upload via Schema Validation 2.0 interface
5. Verify in Security > Events
```

**Fallthrough rule** (catch-all unknown endpoints):
```
Security > API Shield > Settings > Fallthrough > Use Template
- Select hostnames
- Create rule with cf.api_gateway.fallthrough_triggered
- Action: Log (discover) or Block (strict)
```

**Body inspection:** Supports `application/json`, `*/*`, `application/*`. Disable origin MIME sniffing to prevent bypasses.

## JWT Validation

**Setup token config:**
```
Security > API Shield > Settings > JWT Settings > Add configuration
- Name: "Auth0 JWT Config"
- Location: Header/Cookie + name (e.g., "Authorization")
- JWKS: Paste public keys from IdP
```

**Create validation rule:**
```
Security > API Shield > API Rules > Add rule
- Hostname: api.example.com
- Deselect endpoints to ignore
- Token config: Select config
- Enforce presence: Ignore or Mark as non-compliant
- Action: Log/Block/Challenge
```

**Rate limit by JWT claim:**
```wirefilter
lookup_json_string(http.request.jwt.claims["{config_id}"][0], "sub")
```

**Special cases:**
- Two JWTs, different IdPs: Create 2 configs, select both, "Validate all"
- IdP migration: 2 configs + 2 rules, adjust actions per state
- Bearer prefix: API Shield handles with/without
- Nested claims: Dot notation `user.email`

## Mutual TLS (mTLS)

**Setup:**
```
SSL/TLS > Client Certificates > Create Certificate
- Generate CF-managed CA (all plans)
- Upload custom CA (Enterprise, max 5)
```

**Configure mTLS rule:**
```
Security > API Shield > mTLS
- Select hostname(s)
- Choose certificate(s)
- Action: Block/Log/Challenge
```

**Test:**
```bash
openssl req -x509 -newkey rsa:4096 -keyout client-key.pem -out client-cert.pem -days 365
curl https://api.example.com/endpoint --cert client-cert.pem --key client-key.pem
```

## Session Identifiers

Critical for BOLA Detection, Sequence Mitigation, and analytics. Configure header/cookie that uniquely IDs API users.

**Examples:** JWT sub claim, session token, API key, custom user ID header

**Configure:**
```
Security > API Shield > Settings > Session Identifiers
- Type: Header/Cookie
- Name: "X-User-ID" or "Authorization"
```

## BOLA Detection

Detects Broken Object Level Authorization attacks (enumeration + parameter pollution).

**Enable:**
```
Security > API Shield > Schema Validation > [Select Schema] > BOLA Detection
- Enable detection
- Threshold: Sensitivity level (Low/Medium/High)
- Action: Log or Block
```

**Requirements:**
- Schema Validation 2.0 enabled
- Session identifiers configured
- Minimum traffic: 1000+ requests/day per endpoint

## Authentication Posture

Identifies unprotected or inconsistently protected endpoints.

**View report:**
```
Security > API Shield > Authentication Posture
- Shows endpoints lacking JWT/mTLS
- Highlights mixed authentication patterns
```

**Remediate:**
1. Review flagged endpoints
2. Add JWT validation rules
3. Configure mTLS for sensitive endpoints
4. Monitor posture score

## Volumetric Abuse + GraphQL

**Volumetric Abuse Detection:**
`Security > API Shield > Settings > Volumetric Abuse Detection`
- Enable per-endpoint monitoring, set thresholds, action: Log | Challenge | Block

**GraphQL Protection:**
`Security > API Shield > Settings > GraphQL Protection`
- Max query depth: 10, max size: 100KB, block introspection (production)

## Terraform

```hcl
# Session identifier
resource "cloudflare_api_shield" "main" {
  zone_id = var.zone_id
  auth_id_characteristics {
    type = "header"
    name = "Authorization"
  }
}

# Add endpoint
resource "cloudflare_api_shield_operation" "users_get" {
  zone_id  = var.zone_id
  method   = "GET"
  host     = "api.example.com"
  endpoint = "/api/users/{id}"
}

# JWT validation rule
resource "cloudflare_ruleset" "jwt_validation" {
  zone_id = var.zone_id
  name    = "API JWT Validation"
  kind    = "zone"
  phase   = "http_request_firewall_custom"

  rules {
    action = "block"
    expression = "(http.host eq \"api.example.com\" and not is_jwt_valid(http.request.jwt.payload[\"{config_id}\"][0]))"
    description = "Block invalid JWTs"
  }
}
```

## See Also

- [api.md](api.md) - API endpoints and Workers integration
- [patterns.md](patterns.md) - Firewall rules and deployment patterns
- [gotchas.md](gotchas.md) - Troubleshooting and limits

# Gotchas & Troubleshooting

## Common Errors

### "Schema Validation 2.0 not working after migration"

**Cause:** Classic rules still active, conflicting with new system
**Solution:**
1. Delete ALL Classic schema validation rules
2. Clear Cloudflare cache (wait 5 min)
3. Re-upload schema via new Schema Validation 2.0 interface
4. Verify in Security > Events
5. Check action is set (Log/Block)

### "Schema validation blocking valid requests"

**Cause:** Schema too restrictive, missing fields, or incorrect types
**Solution:** 
1. Check Firewall Events for violation details
2. Review schema in Settings
3. Test schema in Swagger Editor
4. Use Log mode to validate before blocking
5. Update schema with correct specifications
6. Ensure Schema Validation 2.0 (not Classic)

### "JWT validation failing"

**Cause:** JWKS mismatch with IdP, expired token, wrong header/cookie name, or clock skew
**Solution:** 
1. Verify JWKS matches IdP configuration
2. Check token `exp` claim is valid
3. Confirm header/cookie name matches config
4. Test token at jwt.io
5. Account for clock skew (±5 min tolerance)
6. Use modern syntax: `is_jwt_valid(http.request.jwt.payload["{config_id}"][0])`

### "BOLA detection false positives"

**Cause:** Legitimate sequential access patterns, bulk operations, or sensitivity too high
**Solution:**
1. Review BOLA events in Security > Events
2. Lower sensitivity threshold (High → Medium → Low)
3. Exclude legitimate bulk operations from detection
4. Ensure session identifiers uniquely identify users
5. Verify minimum traffic requirements met (1000+ req/day)

### "Risk labels not appearing in firewall rules"

**Cause:** Feature not enabled, insufficient traffic, or missing session identifiers
**Solution:**
1. Verify Schema Validation 2.0 enabled
2. Enable BOLA Detection in schema settings
3. Configure session identifiers (required for BOLA)
4. Wait 24-48h for ML model training
5. Check minimum traffic thresholds met

### "Endpoint discovery not finding APIs"

**Cause:** Insufficient traffic (<500 reqs/10d), non-2xx responses, Worker direct requests, or incorrect session ID config
**Solution:** Ensure 500+ requests in 10 days, 2xx responses from edge (not Workers direct), configure session IDs correctly. ML updates daily.

### "Sequence detection false positives"

**Cause:** Lookback window issues, non-unique session IDs, or model sensitivity
**Solution:** 
1. Review lookback settings (10 reqs to managed endpoints, 10min window)
2. Ensure session ID uniqueness per user (not shared tokens)
3. Adjust positive/negative model balance
4. Exclude legitimate workflows from detection

### "GraphQL protection blocking valid queries"

**Cause:** Query depth/size limits too restrictive, complex but legitimate queries
**Solution:**
1. Review blocked query patterns in Security > Events
2. Increase max_depth (default: 10) if needed
3. Increase max_size (default: 100KB) for complex queries
4. Whitelist specific query signatures
5. Use Log mode to tune before blocking

### "Token invalid"

**Cause:** Configuration error, JWKS mismatch, or expired token
**Solution:** Verify config matches IdP, update JWKS, check token expiration

### "Schema violation"

**Cause:** Missing required fields, wrong data types, or spec mismatch
**Solution:** Review schema against actual requests, ensure all required fields present, validate types match spec

### "Fallthrough"

**Cause:** Unknown endpoint or pattern mismatch
**Solution:** Update schema with all endpoints, check path pattern matching

### "mTLS failed"

**Cause:** Certificate untrusted/expired or wrong CA
**Solution:** Verify cert chain, check expiration, confirm correct CA uploaded

## Limits (2026)

| Resource/Limit | Value | Notes |
|----------------|-------|-------|
| OpenAPI version | v3.0.x only | No external refs, must be valid |
| Schema operations | 10K (Enterprise) | Contact for higher limits |
| JWT validation sources | Headers/cookies only | No query params/body |
| Endpoint discovery | 500+ reqs/10d | Minimum for ML model |
| Path normalization | Automatic | `/profile/238` → `/profile/{var1}` |
| Schema parameters | No `content` field | No object param validation |
| BOLA detection | 1000+ reqs/day/endpoint | Per-endpoint minimum |
| Session ID uniqueness | Required | BOLA/Sequence need unique IDs |
| GraphQL max depth | 1-50 | Default: 10 |
| GraphQL max size | 1KB-1MB | Default: 100KB |
| JWT claim nesting | 10 levels max | Use dot notation |
| mTLS CA certificates | 5 custom max | CF-managed unlimited |
| Schema upload size | 5MB max | Compressed OpenAPI spec |
| Volumetric abuse baseline | 7 days training | Initial ML period |
| Auth Posture refresh | Daily | Updated nightly |

## See Also

- [configuration.md](configuration.md) - Setup guides to avoid common issues
- [patterns.md](patterns.md) - Best practices and progressive rollout
- [API Shield Docs](https://developers.cloudflare.com/api-shield/)

# Patterns & Use Cases

## Protect API with Schema + JWT

```bash
# 1. Upload OpenAPI schema
POST /zones/{zone_id}/api_gateway/user_schemas

# 2. Configure JWT validation
POST /zones/{zone_id}/api_gateway/token_validation
{
  "name": "Auth0",
  "location": {"header": "Authorization"},
  "jwks": "{...}"
}

# 3. Create JWT rule
POST /zones/{zone_id}/api_gateway/jwt_validation_rules

# 4. Set schema validation action
PUT /zones/{zone_id}/api_gateway/settings/schema_validation
{"validation_default_mitigation_action": "block"}
```

## Progressive Rollout

```
1. Log mode: Observe false positives
   - Schema: Action = Log
   - JWT: Action = Log

2. Block subset: Protect critical endpoints
   - Change specific endpoint actions to Block
   - Monitor firewall events

3. Full enforcement: Block all violations
   - Change default action to Block
   - Handle fallthrough with custom rule
```

## BOLA Detection

### Enumeration Detection
Detects sequential resource access (e.g., `/users/1`, `/users/2`, `/users/3`).

```javascript
// Block BOLA enumeration attempts
(cf.api_gateway.cf-risk-bola-enumeration and http.host eq "api.example.com")
// Action: Block or Challenge
```

### Parameter Pollution
Detects duplicate/excessive parameters in requests.

```javascript
// Block parameter pollution
(cf.api_gateway.cf-risk-bola-pollution and http.host eq "api.example.com")
// Action: Block
```

### Combined BOLA Protection
```javascript
// Comprehensive BOLA rule
(cf.api_gateway.cf-risk-bola-enumeration or cf.api_gateway.cf-risk-bola-pollution)
and http.host eq "api.example.com"
// Action: Block
```

## Authentication Posture

### Detect Missing Auth
```javascript
// Log endpoints lacking authentication
(cf.api_gateway.cf-risk-missing-auth and http.host eq "api.example.com")
// Action: Log (for audit)
```

### Detect Mixed Auth
```javascript
// Alert on inconsistent auth patterns
(cf.api_gateway.cf-risk-mixed-auth and http.host eq "api.example.com")
// Action: Log (review required)
```

## Fallthrough Detection (Shadow APIs)

```javascript
// WAF Custom Rule
(cf.api_gateway.fallthrough_triggered and http.host eq "api.example.com")
// Action: Log (discover unknown) or Block (strict)
```

## Rate Limiting by User

```javascript
// Rate Limiting Rule (modern syntax)
(http.host eq "api.example.com" and
 is_jwt_valid(http.request.jwt.payload["{config_id}"][0]))

// Rate: 100 req/60s
// Counting expression: lookup_json_string(http.request.jwt.payload["{config_id}"][0], "sub")
```

## Volumetric Abuse Response

```javascript
// Detect abnormal traffic spikes
(cf.api_gateway.volumetric_abuse_detected and http.host eq "api.example.com")
// Action: Challenge or Rate Limit

// Combined with rate limiting
(cf.api_gateway.volumetric_abuse_detected or
 cf.threat_score gt 50) and http.host eq "api.example.com"
// Action: JS Challenge
```

## GraphQL Protection

```javascript
// Block oversized queries
(http.request.uri.path eq "/graphql" and
 cf.api_gateway.graphql_query_size gt 100000)
// Action: Block

// Block deep nested queries
(http.request.uri.path eq "/graphql" and
 cf.api_gateway.graphql_query_depth gt 10)
// Action: Block
```

## Architecture Patterns

**Public API:** Discovery + Schema Validation 2.0 + JWT + Rate Limiting + Bot Management  
**Partner API:** mTLS + Schema Validation + Sequence Mitigation  
**Internal API:** Discovery + Schema Learning + Auth Posture

## OWASP API Security Top 10 Mapping (2026)

| OWASP Issue | API Shield Solutions |
|-------------|---------------------|
| API1:2023 Broken Object Level Authorization | **BOLA Detection** (enumeration + pollution), Sequence mitigation, Schema, JWT, Rate Limiting |
| API2:2023 Broken Authentication | **Auth Posture**, mTLS, JWT validation, Bot Management |
| API3:2023 Broken Object Property Auth | Schema validation, JWT validation |
| API4:2023 Unrestricted Resource Access | Rate Limiting, **Volumetric Abuse Detection**, **GraphQL Protection**, Bot Management |
| API5:2023 Broken Function Level Auth | Schema validation, JWT validation, Auth Posture |
| API6:2023 Unrestricted Business Flows | Sequence mitigation, Bot Management |
| API7:2023 SSRF | Schema validation, WAF managed rules |
| API8:2023 Security Misconfiguration | **Schema Validation 2.0**, Auth Posture, WAF rules |
| API9:2023 Improper Inventory Management | **API Discovery**, Schema learning, Auth Posture |
| API10:2023 Unsafe API Consumption | JWT validation, Schema validation, WAF managed |

## Monitoring

**Security Events:** `Security > Events` → Filter: Action = block, Service = API Shield  
**Firewall Analytics:** `Analytics > Security` → Filter by `cf.api_gateway.*` fields  
**Logpush fields:** APIGatewayAuthIDPresent, APIGatewayRequestViolatesSchema, APIGatewayFallthroughDetected, JWTValidationResult

## Availability (2026)

| Feature | Availability | Notes |
|---------|-------------|-------|
| mTLS (CF-managed CA) | All plans | Self-service |
| Endpoint Management | All plans | Limited operations |
| Schema Validation 2.0 | All plans | Limited operations |
| API Discovery | Enterprise | 10K+ ops |
| JWT Validation | Enterprise add-on | Full validation |
| BOLA Detection | Enterprise add-on | Requires session IDs |
| Auth Posture | Enterprise add-on | Security audit |
| Volumetric Abuse Detection | Enterprise add-on | Traffic analysis |
| GraphQL Protection | Enterprise add-on | Query limits |
| Sequence Mitigation | Enterprise (beta) | Contact team |
| Full Suite | Enterprise add-on | All features |

**Enterprise limits:** 10K operations (contact for higher). Preview access available for non-contract evaluation.

## See Also

- [configuration.md](configuration.md) - Setup all features before creating rules
- [api.md](api.md) - Firewall field reference and API endpoints
- [gotchas.md](gotchas.md) - Common issues and limits

# Cloudflare API Shield Reference

Expert guidance for API Shield - comprehensive API security suite for discovery, protection, and monitoring.

## Reading Order

| Task | Files to Read |
|------|---------------|
| Initial setup | README → configuration.md |
| Implement JWT validation | configuration.md → api.md |
| Add schema validation | configuration.md → patterns.md |
| Detect API attacks | patterns.md → api.md |
| Debug issues | gotchas.md |

## Feature Selection

What protection do you need?

```
├─ Validate request/response structure → Schema Validation 2.0 (configuration.md)
├─ Verify auth tokens → JWT Validation (configuration.md)
├─ Client certificates → mTLS (configuration.md)
├─ Detect BOLA attacks → BOLA Detection (patterns.md)
├─ Track auth coverage → Auth Posture (patterns.md)
├─ Stop volumetric abuse → Abuse Detection (patterns.md)
└─ Discover shadow APIs → API Discovery (api.md)
```

## In This Reference

- **[configuration.md](configuration.md)** - Setup, session identifiers, rules, token/mTLS configs
- **[api.md](api.md)** - Endpoint management, discovery, validation APIs, GraphQL operations
- **[patterns.md](patterns.md)** - Common patterns, progressive rollout, OWASP mappings, workflows
- **[gotchas.md](gotchas.md)** - Troubleshooting, false positives, performance, best practices

## Quick Start

API Shield: Enterprise-grade API security (Discovery, Schema Validation 2.0, JWT, mTLS, BOLA Detection, Auth Posture). Available as Enterprise add-on with preview access.

## See Also

- [API Shield Docs](https://developers.cloudflare.com/api-shield/)
- [API Reference](https://developers.cloudflare.com/api/resources/api_gateway/)
- [OWASP API Security Top 10](https://owasp.org/www-project-api-security/)

## API Reference

**Note on Smart Shield:** Argo Smart Routing is being integrated into Cloudflare's Smart Shield product. API endpoints remain stable; existing integrations continue to work without changes.

### Base Endpoint
```
https://api.cloudflare.com/client/v4
```

### Authentication
Use API tokens with Zone:Argo Smart Routing:Edit permissions:

```bash
# Headers required
X-Auth-Email: user@example.com
Authorization: Bearer YOUR_API_TOKEN
```

### Get Argo Smart Routing Status

**Endpoint:** `GET /zones/{zone_id}/argo/smart_routing`

**Description:** Retrieves current Argo Smart Routing enablement status.

**cURL Example:**
```bash
curl -X GET "https://api.cloudflare.com/client/v4/zones/{zone_id}/argo/smart_routing" \
  -H "Authorization: Bearer YOUR_API_TOKEN" \
  -H "Content-Type: application/json"
```

**Response:**
```json
{
  "result": {
    "id": "smart_routing",
    "value": "on",
    "editable": true,
    "modified_on": "2024-01-11T12:00:00Z"
  },
  "success": true,
  "errors": [],
  "messages": []
}
```

**TypeScript SDK Example:**
```typescript
import Cloudflare from 'cloudflare';

const client = new Cloudflare({
  apiToken: process.env.CLOUDFLARE_API_TOKEN
});

const status = await client.argo.smartRouting.get({ zone_id: 'your-zone-id' });
console.log(`Argo status: ${status.value}, editable: ${status.editable}`);
```

**Python SDK Example:**
```python
from cloudflare import Cloudflare

client = Cloudflare(api_token=os.environ.get('CLOUDFLARE_API_TOKEN'))

status = client.argo.smart_routing.get(zone_id='your-zone-id')
print(f"Argo status: {status.value}, editable: {status.editable}")
```

### Update Argo Smart Routing Status

**Endpoint:** `PATCH /zones/{zone_id}/argo/smart_routing`

**Description:** Enable or disable Argo Smart Routing for a zone.

**Request Body:**
```json
{
  "value": "on"  // or "off"
}
```

**cURL Example:**
```bash
curl -X PATCH "https://api.cloudflare.com/client/v4/zones/{zone_id}/argo/smart_routing" \
  -H "Authorization: Bearer YOUR_API_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"value": "on"}'
```

**TypeScript SDK Example:**
```typescript
const result = await client.argo.smartRouting.edit({
  zone_id: 'your-zone-id',
  value: 'on',
});
console.log(`Updated: ${result.value} at ${result.modified_on}`);
```

**Python SDK Example:**
```python
result = client.argo.smart_routing.edit(
    zone_id='your-zone-id',
    value='on'
)
print(f"Updated: {result.value} at {result.modified_on}")
```

## Checking Editability Before Updates

**Critical:** Always check the `editable` field before attempting to enable/disable Argo. When `editable: false`, the zone has restrictions (billing not configured, insufficient permissions, or plan limitations).

**Pattern:**
```typescript
async function safelyEnableArgo(client: Cloudflare, zoneId: string): Promise<boolean> {
  const status = await client.argo.smartRouting.get({ zone_id: zoneId });
  
  if (!status.editable) {
    console.error('Cannot modify Argo: editable=false (check billing/permissions)');
    return false;
  }
  
  if (status.value === 'on') {
    console.log('Argo already enabled');
    return true;
  }
  
  await client.argo.smartRouting.edit({ zone_id: zoneId, value: 'on' });
  console.log('Argo enabled successfully');
  return true;
}
```

**Python Pattern:**
```python
def safely_enable_argo(client: Cloudflare, zone_id: str) -> bool:
    status = client.argo.smart_routing.get(zone_id=zone_id)
    
    if not status.editable:
        print('Cannot modify Argo: editable=false (check billing/permissions)')
        return False
    
    if status.value == 'on':
        print('Argo already enabled')
        return True
    
    client.argo.smart_routing.edit(zone_id=zone_id, value='on')
    print('Argo enabled successfully')
    return True
```

## Error Handling

The TypeScript SDK provides typed error classes for robust error handling:

```typescript
import Cloudflare from 'cloudflare';
import { APIError, APIConnectionError, RateLimitError } from 'cloudflare';

async function enableArgoWithErrorHandling(client: Cloudflare, zoneId: string) {
  try {
    const result = await client.argo.smartRouting.edit({
      zone_id: zoneId,
      value: 'on',
    });
    return result;
  } catch (error) {
    if (error instanceof RateLimitError) {
      console.error('Rate limited. Retry after:', error.response?.headers.get('retry-after'));
      // Implement exponential backoff
    } else if (error instanceof APIError) {
      console.error('API error:', error.status, error.message);
      if (error.status === 403) {
        console.error('Permission denied - check API token scopes');
      } else if (error.status === 400) {
        console.error('Bad request - verify zone_id and payload');
      }
    } else if (error instanceof APIConnectionError) {
      console.error('Connection failed:', error.message);
      // Retry with exponential backoff
    } else {
      console.error('Unexpected error:', error);
    }
    throw error;
  }
}
```

**Python Error Handling:**
```python
from cloudflare import Cloudflare, APIError, RateLimitError

def enable_argo_with_error_handling(client: Cloudflare, zone_id: str):
    try:
        result = client.argo.smart_routing.edit(zone_id=zone_id, value='on')
        return result
    except RateLimitError as e:
        print(f"Rate limited. Retry after: {e.response.headers.get('retry-after')}")
        raise
    except APIError as e:
        print(f"API error: {e.status} - {e.message}")
        if e.status == 403:
            print('Permission denied - check API token scopes')
        elif e.status == 400:
            print('Bad request - verify zone_id and payload')
        raise
    except Exception as e:
        print(f"Unexpected error: {e}")
        raise
```

## Response Schema

All Argo Smart Routing API responses follow this structure:

```typescript
interface ArgoSmartRoutingResponse {
  result: {
    id: 'smart_routing';
    value: 'on' | 'off';
    editable: boolean;
    modified_on: string; // ISO 8601 timestamp
  };
  success: boolean;
  errors: Array<{
    code: number;
    message: string;
  }>;
  messages: Array<string>;
}
```

## Key Response Fields

| Field | Type | Description |
|-------|------|-------------|
| `value` | `"on" \| "off"` | Current enablement status |
| `editable` | `boolean` | Whether changes are allowed (check before PATCH) |
| `modified_on` | `string` | ISO timestamp of last modification |
| `success` | `boolean` | Whether request succeeded |
| `errors` | `Array` | Error details if `success: false`

## Configuration Management

**Note on Smart Shield Evolution:** Argo Smart Routing is being integrated into Smart Shield. Configuration methods below remain valid; Terraform and IaC patterns unchanged.

### Infrastructure as Code (Terraform)

```hcl
# terraform/argo.tf
# Note: Use Cloudflare Terraform provider

resource "cloudflare_argo" "example" {
  zone_id        = var.zone_id
  smart_routing  = "on"
  tiered_caching = "on"
}

variable "zone_id" {
  description = "Cloudflare Zone ID"
  type        = string
}

output "argo_enabled" {
  value       = cloudflare_argo.example.smart_routing
  description = "Argo Smart Routing status"
}
```

### Environment-Based Configuration

```typescript
// config/argo.ts
interface ArgoEnvironmentConfig {
  enabled: boolean;
  tieredCache: boolean;
  monitoring: {
    usageAlerts: boolean;
    threshold: number;
  };
}

const configs: Record<string, ArgoEnvironmentConfig> = {
  production: {
    enabled: true,
    tieredCache: true,
    monitoring: {
      usageAlerts: true,
      threshold: 1000, // GB
    },
  },
  staging: {
    enabled: true,
    tieredCache: false,
    monitoring: {
      usageAlerts: false,
      threshold: 100, // GB
    },
  },
  development: {
    enabled: false,
    tieredCache: false,
    monitoring: {
      usageAlerts: false,
      threshold: 0,
    },
  },
};

export function getArgoConfig(env: string): ArgoEnvironmentConfig {
  return configs[env] || configs.development;
}
```

### Pulumi Configuration

```typescript
// pulumi/argo.ts
import * as cloudflare from '@pulumi/cloudflare';

const zone = new cloudflare.Zone('example-zone', {
  zone: 'example.com',
  plan: 'enterprise',
});

const argoSettings = new cloudflare.Argo('argo-config', {
  zoneId: zone.id,
  smartRouting: 'on',
  tieredCaching: 'on',
});

export const argoEnabled = argoSettings.smartRouting;
export const zoneId = zone.id;
```

## Billing Configuration

Before enabling Argo Smart Routing, ensure billing is configured for the account:

**Prerequisites:**
1. Valid payment method on file
2. Enterprise or higher plan
3. Zone must have billing enabled

**Check Billing Status via Dashboard:**
1. Navigate to Account → Billing
2. Verify payment method configured
3. Check zone subscription status

**Note:** Attempting to enable Argo without billing configured will result in `editable: false` in API responses.

## Environment Variable Setup

**Required Environment Variables:**
```bash
# .env
CLOUDFLARE_API_TOKEN=your_api_token_here
CLOUDFLARE_ZONE_ID=your_zone_id_here
CLOUDFLARE_ACCOUNT_ID=your_account_id_here

# Optional
ARGO_ENABLED=true
ARGO_TIERED_CACHE=true
```

**TypeScript Configuration Loader:**
```typescript
// config/env.ts
import { z } from 'zod';

const envSchema = z.object({
  CLOUDFLARE_API_TOKEN: z.string().min(1),
  CLOUDFLARE_ZONE_ID: z.string().min(1),
  CLOUDFLARE_ACCOUNT_ID: z.string().min(1),
  ARGO_ENABLED: z.string().optional().default('false'),
  ARGO_TIERED_CACHE: z.string().optional().default('false'),
});

export const env = envSchema.parse(process.env);

export const argoConfig = {
  enabled: env.ARGO_ENABLED === 'true',
  tieredCache: env.ARGO_TIERED_CACHE === 'true',
};
```

## CI/CD Integration

**GitHub Actions Example:**
```yaml
# .github/workflows/deploy-argo.yml
name: Deploy Argo Configuration

on:
  push:
    branches: [main]
    paths:
      - 'terraform/argo.tf'

jobs:
  deploy:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v3
      
      - name: Setup Terraform
        uses: hashicorp/setup-terraform@v2
        
      - name: Terraform Init
        run: terraform init
        working-directory: ./terraform
        
      - name: Terraform Apply
        run: terraform apply -auto-approve
        working-directory: ./terraform
        env:
          CLOUDFLARE_API_TOKEN: ${{ secrets.CLOUDFLARE_API_TOKEN }}
          TF_VAR_zone_id: ${{ secrets.CLOUDFLARE_ZONE_ID }}
```

## Enterprise Preview Program

For early access to Argo Smart Routing features and Smart Shield integration:

**Eligibility:**
- Enterprise plan customers
- Active Cloudflare support contract
- Production traffic >100GB/month

**How to Join:**
1. Contact Cloudflare account team or support
2. Request Argo/Smart Shield preview access
3. Receive preview zone configuration

**Preview Features:**
- Enhanced analytics and reporting
- Smart Shield DDoS integration
- Advanced routing policies
- Priority support for routing issues

## Best Practices Summary

**Smart Shield Note:** Argo Smart Routing evolving into Smart Shield. Best practices below remain applicable; monitor Cloudflare changelog for Smart Shield updates.

1. **Always check editability** before attempting to enable/disable Argo
2. **Set up billing notifications** to avoid unexpected costs
3. **Combine with Tiered Cache** for maximum performance benefit
4. **Use in production only** - disable for dev/staging to control costs
5. **Monitor analytics** - require 500+ requests in 48h for detailed metrics
6. **Handle errors gracefully** - check for billing, permissions, zone compatibility
7. **Test configuration changes** in staging before production
8. **Use TypeScript SDK** for type safety and better developer experience
9. **Implement retry logic** for API calls in production systems
10. **Document zone-specific settings** for team visibility

## Common Errors

### "Argo unavailable"

**Problem:** API returns error "Argo Smart Routing is unavailable for this zone"

**Cause:** Zone not eligible or billing not set up

**Solution:**
1. Verify zone has Enterprise or higher plan
2. Check billing is configured in Account → Billing
3. Ensure payment method is valid and current
4. Contact Cloudflare support if eligibility unclear

### "Cannot enable/disable"

**Problem:** API call succeeds but status remains unchanged, or `editable: false` in GET response

**Cause:** Insufficient permissions or zone restrictions

**Solution:**
1. Check API token has `Zone:Argo Smart Routing:Edit` permission
2. Verify `editable: true` in GET response before attempting PATCH
3. If `editable: false`, check:
   - Billing configured for account
   - Zone plan includes Argo (Enterprise+)
   - No active zone holds or suspensions
   - API token has correct scopes

### `editable: false` Error

**Problem:** GET request returns `"editable": false`, preventing enable/disable

**Cause:** Zone-level restrictions from billing, plan, or permissions

**Solution Pattern:**
```typescript
const status = await client.argo.smartRouting.get({ zone_id: zoneId });

if (!status.editable) {
  // Don't attempt to modify - will fail
  console.error('Cannot modify Argo settings:');
  console.error('- Check billing is configured');
  console.error('- Verify zone has Enterprise+ plan');
  console.error('- Confirm API token has Edit permission');
  throw new Error('Argo is not editable for this zone');
}

// Safe to proceed with enable/disable
await client.argo.smartRouting.edit({ zone_id: zoneId, value: 'on' });
```

### Rate Limiting

**Problem:** `429 Too Many Requests` error from API

**Cause:** Exceeded API rate limits (typically 1200 requests per 5 minutes)

**Solution:**
```typescript
import { RateLimitError } from 'cloudflare';

try {
  await client.argo.smartRouting.edit({ zone_id: zoneId, value: 'on' });
} catch (error) {
  if (error instanceof RateLimitError) {
    const retryAfter = error.response?.headers.get('retry-after');
    console.log(`Rate limited. Retry after ${retryAfter} seconds`);
    
    // Implement exponential backoff
    await new Promise(resolve => setTimeout(resolve, (retryAfter || 60) * 1000));
    // Retry request
  }
}
```

## Limits

| Resource/Limit | Value | Notes |
|----------------|-------|-------|
| Min requests for analytics | 500 in 48h | For detailed metrics via GraphQL |
| Zones supported | Enterprise+ | Check zone plan in dashboard |
| Billing requirement | Must be configured | Before enabling; verify payment method |
| API rate limit | 1200 req / 5 min | Per API token across all endpoints |
| Spectrum apps | No hard limit | Each app can enable Argo independently |
| Traffic counting | Proxied only | Only orange-clouded DNS records count |
| DDoS/WAF exemption | Yes | Mitigated traffic excluded from billing |
| Analytics latency | 1-5 minutes | Real-time metrics not available |

## Additional Resources

- [Official Argo Smart Routing Docs](https://developers.cloudflare.com/argo-smart-routing/)
- [Cloudflare Smart Shield](https://developers.cloudflare.com/smart-shield/)
- [API Authentication](https://developers.cloudflare.com/fundamentals/api/get-started/create-token/)
- [Cloudflare TypeScript SDK](https://github.com/cloudflare/cloudflare-typescript)
- [Cloudflare Python SDK](https://github.com/cloudflare/cloudflare-python)

# Integration Patterns

## Enable Argo + Tiered Cache

```typescript
async function enableOptimalPerformance(client: Cloudflare, zoneId: string) {
  await Promise.all([
    client.argo.smartRouting.edit({ zone_id: zoneId, value: 'on' }),
    client.argo.tieredCaching.edit({ zone_id: zoneId, value: 'on' }),
  ]);
}
```

**Flow:** Visitor → Edge (Lower-Tier) → [Cache Miss] → Upper-Tier → [Cache Miss + Argo] → Origin

**Impact:** Argo ~30% latency reduction + Tiered Cache 50-80% origin offload

## Usage Analytics (GraphQL)

```graphql
query ArgoAnalytics($zoneTag: string!) {
  viewer {
    zones(filter: { zoneTag: $zoneTag }) {
      httpRequestsAdaptiveGroups(limit: 1000) {
        sum { argoBytes, bytes }
      }
    }
  }
}
```

**Billing:** ~$0.10/GB. DDoS-mitigated and WAF-blocked traffic NOT charged.

## Spectrum TCP Integration

Enable Argo for non-HTTP traffic (databases, game servers, IoT):

```typescript
// Update existing app
await client.spectrum.apps.update(appId, { zone_id: zoneId, argo_smart_routing: true });

// Create new app with Argo
await client.spectrum.apps.create({
  zone_id: zoneId,
  dns: { type: 'CNAME', name: 'tcp.example.com' },
  origin_direct: ['tcp://origin.example.com:3306'],
  protocol: 'tcp/3306',
  argo_smart_routing: true,
});
```

**Use cases:** MySQL/PostgreSQL (3306/5432), game servers, MQTT (1883), SSH (22)

## Pre-Flight Validation

```typescript
async function validateArgoEligibility(client: Cloudflare, zoneId: string) {
  const status = await client.argo.smartRouting.get({ zone_id: zoneId });
  const zone = await client.zones.get({ zone_id: zoneId });
  
  const issues: string[] = [];
  if (!status.editable) issues.push('Zone not editable');
  if (['free', 'pro'].includes(zone.plan.legacy_id)) issues.push('Requires Business+ plan');
  if (zone.status !== 'active') issues.push('Zone not active');
  
  return { canEnable: issues.length === 0, issues };
}
```

## Post-Enable Verification

```typescript
async function verifyArgoEnabled(client: Cloudflare, zoneId: string): Promise<boolean> {
  await new Promise(r => setTimeout(r, 2000)); // Wait for propagation
  const status = await client.argo.smartRouting.get({ zone_id: zoneId });
  return status.value === 'on';
}
```

## Full Setup Pattern

```typescript
async function setupArgo(client: Cloudflare, zoneId: string) {
  // 1. Validate
  const { canEnable, issues } = await validateArgoEligibility(client, zoneId);
  if (!canEnable) throw new Error(issues.join(', '));
  
  // 2. Enable both features
  await Promise.all([
    client.argo.smartRouting.edit({ zone_id: zoneId, value: 'on' }),
    client.argo.tieredCaching.edit({ zone_id: zoneId, value: 'on' }),
  ]);
  
  // 3. Verify
  const [argo, cache] = await Promise.all([
    client.argo.smartRouting.get({ zone_id: zoneId }),
    client.argo.tieredCaching.get({ zone_id: zoneId }),
  ]);
  
  return { argo: argo.value === 'on', tieredCache: cache.value === 'on' };
}
```

**When to combine:** High-traffic sites (>1TB/mo), global users, cacheable content.

# Cloudflare Argo Smart Routing Skill Reference

## Overview

Cloudflare Argo Smart Routing is a performance optimization service that detects real-time network issues and routes web traffic across the most efficient network path. It continuously monitors network conditions and intelligently routes traffic through the fastest, most reliable routes in Cloudflare's network.

**Note on Smart Shield:** Argo Smart Routing is being integrated into Cloudflare's Smart Shield product for enhanced DDoS protection and performance. Existing Argo customers maintain full functionality with gradual migration to Smart Shield features.

## Quick Start

### Enable via cURL
```bash
curl -X PATCH "https://api.cloudflare.com/client/v4/zones/{zone_id}/argo/smart_routing" \
  -H "Authorization: Bearer YOUR_API_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"value": "on"}'
```

### Enable via TypeScript SDK
```typescript
import Cloudflare from 'cloudflare';

const client = new Cloudflare({ apiToken: process.env.CLOUDFLARE_API_TOKEN });

const result = await client.argo.smartRouting.edit({
  zone_id: 'your-zone-id',
  value: 'on',
});

console.log(`Argo enabled: ${result.value}`);
```

## Core Concepts

### What It Does
- **Intelligent routing**: Detects congestion, outages, packet loss in real-time
- **Global optimization**: Routes across 300+ Cloudflare data centers
- **Automatic failover**: Switches paths when issues detected (typically <1s)
- **Works with existing setup**: No origin changes required

### Billing Model
- Usage-based: Charged per GB of traffic (excluding DDoS/WAF mitigated traffic)
- Requires billing configuration before enabling
- Available on Enterprise+ plans (check zone eligibility)

### When to Use
- **High-traffic production sites** with global user base
- **Latency-sensitive applications** (APIs, real-time services)
- **Sites behind Cloudflare proxy** (orange-clouded DNS records)
- **Combined with Tiered Cache** for maximum performance gains

### When NOT to Use
- Development/staging environments (cost control)
- Low-traffic sites (<1TB/month) where cost may exceed benefit
- Sites with primarily single-region traffic

## Should I Enable Argo?

| Your Situation | Recommendation |
|----------------|----------------|
| Global production app, >1TB/month traffic | ✅ Enable - likely ROI positive |
| Enterprise plan, latency-critical APIs | ✅ Enable - performance matters |
| Regional site, <100GB/month traffic | ⚠️ Evaluate - cost may not justify |
| Development/staging environment | ❌ Disable - use in production only |
| Not yet configured billing | ❌ Configure billing first |

## Reading Order by Task

| Your Goal | Start With | Then Read |
|-----------|------------|-----------|
| Enable Argo for first time | Quick Start above → [configuration.md](configuration.md) | [gotchas.md](gotchas.md) |
| Use TypeScript/Python SDK | [api.md](api.md) | [patterns.md](patterns.md) |
| Terraform/IaC setup | [configuration.md](configuration.md) | - |
| Enable for Spectrum TCP app | [patterns.md](patterns.md) → Spectrum section | [api.md](api.md) |
| Troubleshoot enablement issue | [gotchas.md](gotchas.md) | [api.md](api.md) |
| Manage billing/usage | [patterns.md](patterns.md) → Billing section | [gotchas.md](gotchas.md) |

## In This Reference

- **[api.md](api.md)** - API endpoints, SDK methods, error handling, Python/TypeScript examples
- **[configuration.md](configuration.md)** - Terraform setup, environment config, billing configuration
- **[patterns.md](patterns.md)** - Tiered Cache integration, Spectrum TCP apps, billing management, validation patterns
- **[gotchas.md](gotchas.md)** - Common errors, permission issues, limits, best practices

## See Also

- [Cloudflare Argo Smart Routing Docs](https://developers.cloudflare.com/argo-smart-routing/)
- [Cloudflare Smart Shield](https://developers.cloudflare.com/smart-shield/)
- [Spectrum Documentation](https://developers.cloudflare.com/spectrum/)
- [Tiered Cache](https://developers.cloudflare.com/cache/how-to/tiered-cache/)

# Artifacts API Reference

Use Artifacts through the **Workers binding**, the **REST control plane**, and **Git-compatible remotes**.

**Prefer retrieval** for exact request and response details. Verify current behavior at `https://developers.cloudflare.com/artifacts/` before relying on specific auth flows, route details, or generated binding types.

## Workers Binding

Artifacts exposes a Worker binding on `env.ARTIFACTS`.

### Namespace Methods

| Method | Use For |
|--------|---------|
| `create(name, opts?)` | Create a repo and receive its initial remote and token |
| `get(name)` | Resolve a repo handle for repo-scoped operations |
| `list(opts?)` | List repos in a namespace |
| `delete(name)` | Delete a repo |

```typescript
const created = await env.ARTIFACTS.create("starter-repo", {
  description: "Repository for automation experiments",
  setDefaultBranch: "main"
});
const repo = await env.ARTIFACTS.get("starter-repo");
const page = await env.ARTIFACTS.list({ limit: 10 });
```

Use the REST API when you need to import a repo from another HTTPS remote.

### Repo Handle Methods

Use a repo handle returned by `get()` or `create()`.

| Method | Use For |
|--------|---------|
| `info()` | Read repo metadata, including the remote URL |
| `createToken(scope?, ttl?)` | Mint a repo-scoped read or write token |
| `listTokens()` | Inspect active tokens |
| `validateToken(token)` | Check whether a token is still valid |
| `revokeToken(tokenOrId)` | Revoke a token by ID or value |
| `fork(name, opts?)` | Fork one repo into another |

```typescript
const repo = await env.ARTIFACTS.get("starter-repo");
if (!repo) throw new Error("Repo not found");

const info = await repo.info();
const token = await repo.createToken("read", 3600);
const forked = await repo.fork("starter-repo-copy", {
  defaultBranchOnly: true
});
```

### Binding Notes

- Current docs describe the runtime binding surface as `create`, `get`, `list`, `delete`, and repo-handle methods like `info`, `createToken`, and `fork`.
- Use `npx wrangler types` in the target project and treat the generated `worker-configuration.d.ts` as the source of truth for that environment.
- If generated types appear to expose `import()` or a different `get()` shape, verify the live docs before depending on those methods.

Verify current runtime behavior in the live docs before depending on methods that are not shown in the Workers binding reference.

## REST API

Artifacts currently documents a namespace-scoped control plane:

```txt
https://artifacts.cloudflare.net/v1/api/namespaces/$ARTIFACTS_NAMESPACE
```

Some deployments also expose an `/edge/v1/api/...` base path. Verify the correct base URL for your environment in the live docs.

Requests to the standard `/v1/api/...` routes use a **gateway JWT** with Bearer authentication.

Returned repo tokens authenticate **Git operations** against the repo `remote`. They do not authenticate REST control-plane requests.

Current docs show the standard Cloudflare v4 response envelope around REST results.

### Repo Routes

| Route | Use For |
|-------|---------|
| `POST /repos` | Create a repo |
| `GET /repos` | List repos |
| `GET /repos/:name` | Read repo metadata and remote |
| `DELETE /repos/:name` | Delete a repo |
| `POST /repos/:name/fork` | Fork a repo |
| `POST /repos/:name/import` | Import a public HTTPS remote |

```bash
curl --request POST "$ARTIFACTS_BASE_URL/repos" \
  --header "Authorization: Bearer $ARTIFACTS_JWT" \
  --header "Content-Type: application/json" \
  --data '{"name":"starter-repo"}'
```

Important current details from the docs draft:
- `POST /repos/:name/import` accepts a full HTTPS remote URL such as GitHub or GitLab.
- Import supports options such as `branch`, `depth`, and `read_only`.
- Repo metadata includes fields such as description, default branch, timestamps, and the Git `remote`.

### Token Routes

| Route | Use For |
|-------|---------|
| `GET /repos/:name/tokens` | List repo tokens |
| `POST /tokens` | Create a token for a repo |
| `DELETE /tokens/:id` | Revoke a token by ID |

Current docs show list-token filtering and pagination by token state. Retrieve the exact query shape from the live docs when you need token audit or cleanup workflows.

Use **read** tokens for clone, fetch, pull, and indexing workflows. Use **write** tokens only when a workflow must push or otherwise mutate a repo.

## Git-Compatible Access

Artifacts returns repo `remote` URLs that work with standard git-over-HTTPS tooling.

Recommended current auth pattern for local workflows:

```bash
git -c http.extraHeader="Authorization: Bearer $ARTIFACTS_TOKEN" clone "$ARTIFACTS_REMOTE" artifacts-clone
```

Use a self-contained Basic-auth remote only for short-lived commands that need credentials embedded in the URL.

`read` tokens support `clone`, `fetch`, and `pull`. `git push` requires a `write` token.

For large repos where startup time matters more than a full clone, Artifacts also documents **ArtifactFS**. Retrieve current details from `https://developers.cloudflare.com/artifacts/` when you need mount-style access.

# Artifacts Configuration

## Worker Binding

Configure the `artifacts` binding in your Wrangler config:

```toml
[[artifacts]]
binding = "ARTIFACTS"
namespace = "default"
```

This exposes Artifacts on `env.ARTIFACTS` inside your Worker.

If you authenticate with `wrangler login`, current docs say Wrangler requests `artifacts:write` by default.

## TypeScript

Regenerate Worker types after adding the binding:

```bash
npx wrangler types
```

Use the generated binding type in your environment definition:

```typescript
interface Env {
  ARTIFACTS: Artifacts;
}
```

Wrangler generates the `Artifacts` type from the binding. Treat the generated `worker-configuration.d.ts` file as the source of truth for your environment.

## Structure Repos for Isolation

Artifacts works best when autonomous work is isolated:
- Create one repo per agent, session, sandbox, or task when work should stay separate.
- Fork from a reviewed baseline instead of copying starter files into every new repo.
- Use branches only when collaborators share the same lifecycle and need to work in one repo.
- Use namespaces to separate environments, teams, or high-rate workloads.

## REST Configuration

For external systems, configure the namespace-scoped base URL and gateway JWT:

```bash
export ARTIFACTS_NAMESPACE="default"
export ARTIFACTS_JWT="<YOUR_GATEWAY_JWT>"
export ARTIFACTS_BASE_URL="https://artifacts.cloudflare.net/v1/api/namespaces/$ARTIFACTS_NAMESPACE"
```

Some environments also expose an `/edge/v1/api/...` base path. Verify the correct host and base path in the live docs for your Artifacts environment.

Use environment variables or your secret manager. Do not hardcode gateway JWTs or repo tokens.

## Repo Tokens

Artifacts workflows usually involve repo-scoped tokens returned by `create()` or minted later through the binding or REST API.

Keep the control plane and data plane separate:
- Use the **Workers binding** or **REST API** with a gateway JWT to create repos and mint tokens.
- Use repo-scoped tokens only for **Git operations** against the returned `remote`.

Recommended handling:
- Mint the narrowest scope you need: `read` or `write`
- Prefer short-lived tokens for handoff between systems
- Revoke tokens that are no longer needed

Verify the current token behavior and auth guidance in `https://developers.cloudflare.com/artifacts/` before building long-lived automation.

## Git Consumers

Artifacts is designed to work with standard git-over-HTTPS clients once you have a repo `remote` and an access token.

Prefer header-based auth for local tooling so the full token stays out of the remote URL:

```bash
git -c http.extraHeader="Authorization: Bearer $ARTIFACTS_TOKEN" clone "$ARTIFACTS_REMOTE" artifacts-clone
```

Use a Basic-auth remote only for short-lived commands that need a self-contained URL.

## Retrieval Checklist

Check the live docs before relying on:
- the current Workers binding surface
- exact token formats
- availability or product status
- route details for import, fork, and token-management flows
- the correct control-plane host or `/edge/v1` base path for your environment
- platform limits or pricing

# Cloudflare Artifacts

Store versioned file trees behind a repo-style interface that works from Workers, the REST API, and Git-compatible tooling.

## Overview

Use **Artifacts** when the thing you need to store is a versioned filesystem tree rather than a single object, key, or SQL row.

Typical Artifacts use cases:
- Git-style repositories
- Per-agent, per-session, or per-task repos
- Build outputs and deployment bundles
- Checkpoints and generated assets
- Shared file trees passed between developer tools and Workers

Artifacts is a good fit when the same content needs to be addressable from **Workers**, the **REST API**, and **Git-compatible clients**.

Artifacts is especially useful for agent and automation workflows where each unit of work should have its own isolated repo and token.

**Prefer retrieval over memory** for current availability, authentication details, route shapes, limits, and pricing. Start at `https://developers.cloudflare.com/artifacts/`.

## When to Use Artifacts

| Need | Use | Why |
|------|-----|-----|
| Versioned file trees such as repos, build outputs, checkpoints, or generated assets | Artifacts | Artifacts stores and shares **versioned filesystem content** |
| A git-compatible workflow with `clone`, `fetch`, `pull`, or `push` | Artifacts | Artifacts exposes **git-over-HTTPS remotes** and repo-scoped tokens |
| The same artifact accessible from Workers, HTTP APIs, and developer tooling | Artifacts | Artifacts is available through a **Workers binding**, **REST API**, and **git-compatible interface** |
| Large files by object key, app config by key, or relational app data | R2, KV, or D1 | Use storage products directly when you need **objects, key-value entries, or SQL rows**, not versioned file trees |

## Recommended Workflow

- Create one repo per agent, session, user workspace, or task when work should stay isolated.
- Fork from a stable baseline when many repos need the same starter files or prompts.
- Use branches only when collaborators share the same lifecycle and need to work in one repo.
- Use namespaces to separate environments, teams, or high-rate workloads.

## Quick Start

**From a Worker:**

```typescript
interface Env {
  ARTIFACTS: Artifacts;
}

const created = await env.ARTIFACTS.create("starter-repo");
// created.remote -> git remote URL
// created.token -> initial repo token
```

**From the REST API:**

Use the namespace-scoped Artifacts base URL plus a gateway JWT. For imports from existing HTTPS remotes, use the REST API rather than the Workers binding.

## Reading Order

| Task | Read |
|------|------|
| Decide whether Artifacts is the right product | README only |
| Create or manage repos from a Worker | README → configuration.md → api.md |
| Integrate Artifacts from an external system | README → api.md |
| Set up agent or sandbox workflows | README → configuration.md |
| Verify exact auth, routes, limits, or pricing | Live docs first: `https://developers.cloudflare.com/artifacts/` |

## In This Reference

- **[api.md](api.md)** - Workers binding methods, REST routes, token and repo operations
- **[configuration.md](configuration.md)** - Wrangler binding shape, Worker typing, REST configuration guidance

## See Also

- [Cloudflare Artifacts Docs](https://developers.cloudflare.com/artifacts/)
- [Artifacts Git Protocol Docs](https://developers.cloudflare.com/artifacts/api/git-protocol/)
- [ArtifactFS Docs](https://developers.cloudflare.com/artifacts/guides/artifact-fs/)
- [Cloudflare Workers Docs](https://developers.cloudflare.com/workers/)
- [Cloudflare Durable Objects Docs](https://developers.cloudflare.com/durable-objects/)
- [Cloudflare R2 Docs](https://developers.cloudflare.com/r2/)
- [Cloudflare D1 Docs](https://developers.cloudflare.com/d1/)

# Bindings API Reference

Fetch the relevant documentation before choosing method signatures, binding types, or type-generation settings.

| Task | Current documentation |
| --- | --- |
| Access bindings in handlers, entrypoint classes, or imported `env` | [Bindings and environment access](https://developers.cloudflare.com/workers/runtime-apis/bindings/) |
| Generate environment and runtime types; configure TypeScript or migrate from `@cloudflare/workers-types` | [Workers TypeScript](https://developers.cloudflare.com/workers/languages/typescript/) |
| Use framework-specific environment access | [Hono on Workers](https://developers.cloudflare.com/workers/framework-guides/web-apps/more-web-frameworks/hono/) (follow the framework's linked documentation) |
| Read, write, delete, and list KV keys | [KV Workers API](https://developers.cloudflare.com/kv/api/) |
| Read, write, delete, and list R2 objects | [R2 Workers API](https://developers.cloudflare.com/r2/api/workers/workers-api-reference/) |
| Prepare and bind SQL statements or execute batches | [D1 Workers Binding API](https://developers.cloudflare.com/d1/worker-api/) |
| Address Durable Objects and call their methods | [Durable Objects API](https://developers.cloudflare.com/durable-objects/api/) |
| Send queue messages | [Queues JavaScript APIs](https://developers.cloudflare.com/queues/configuration/javascript-apis/) |
| Run model inference | [Workers AI bindings](https://developers.cloudflare.com/workers-ai/configuration/bindings/) |
| Call another Worker using HTTP or typed RPC | [Service bindings](https://developers.cloudflare.com/workers/runtime-apis/bindings/service-bindings/) and [RPC TypeScript](https://developers.cloudflare.com/workers/runtime-apis/rpc/typescript/) |
| Find other product binding APIs, including Browser, mTLS, rate limiting, and Workflows | [Current binding catalog](https://developers.cloudflare.com/workers/runtime-apis/bindings/) |

Regenerate types after configuration changes and use the selected environment's configuration. Follow the current TypeScript setup for the project's toolchain instead of hardcoding a generated declaration path or maintaining a handwritten binding interface. Types describe the expected bindings; they do not provision resources or prove the deployed environment is configured correctly.

# Binding Configuration

Confirm the target account, environment, and resource before adding or changing a binding. Keep staging and production resources separate where their data or permissions must be isolated.

| Task | Current documentation |
| --- | --- |
| Configure storage, compute, platform, and service bindings | [Wrangler binding configuration](https://developers.cloudflare.com/workers/wrangler/configuration/#bindings) |
| Create or inspect resources and manage deployments | [Wrangler commands](https://developers.cloudflare.com/workers/wrangler/commands/) (select the product's commands) |
| Configure named environments and non-inheritable bindings | [Wrangler environments](https://developers.cloudflare.com/workers/wrangler/environments/) |
| Set non-sensitive configuration | [Environment variables](https://developers.cloudflare.com/workers/configuration/environment-variables/) |
| Set or rotate credentials in a chosen environment | [Secrets](https://developers.cloudflare.com/workers/configuration/secrets/) |
| Configure text, data, and Wasm modules in existing projects | [Wrangler bundling](https://developers.cloudflare.com/workers/wrangler/bundling/) and [configuration](https://developers.cloudflare.com/workers/wrangler/configuration/) |
| Choose locally simulated resources or remote bindings | [Local development](https://developers.cloudflare.com/workers/local-development/) and [supported bindings per development mode](https://developers.cloudflare.com/workers/local-development/bindings-per-env/) |
| Supply local variables and secrets | [Local environment variables and secrets](https://developers.cloudflare.com/workers/local-development/environment-variables/) |
| Seed or persist local resource data | [Adding local data](https://developers.cloudflare.com/workers/local-development/local-data/) |
| Generate types after configuring bindings | [Workers TypeScript](https://developers.cloudflare.com/workers/languages/typescript/) |

Bindings and variables are not inherited automatically by named environments. Configure the required values for the environment being used, including its secrets. Binding names used in code must match configuration; the target resource's ID or name is a separate value.

Keep credentials out of committed variables and local secret files out of version control. Remote development can access real resources: verify the target rather than assuming that running locally isolates writes.

# Binding Gotchas and Troubleshooting

Start by checking the binding name, selected environment, and actual target resource. Regenerating types alone does not fix a missing runtime binding.

| Symptom or question | What to check |
| --- | --- |
| A binding is undefined or points to unexpected data | [Binding configuration](https://developers.cloudflare.com/workers/wrangler/configuration/#bindings) and [environment inheritance](https://developers.cloudflare.com/workers/wrangler/environments/) |
| TypeScript cannot find a binding or runtime type | [Current type generation and TypeScript setup](https://developers.cloudflare.com/workers/languages/typescript/) |
| Updated credentials are not reflected in a client | [Binding changes and retained global state](https://developers.cloudflare.com/workers/runtime-apis/bindings/#making-changes-to-bindings); verify the [secret's environment and deployment](https://developers.cloudflare.com/workers/configuration/secrets/) |
| Binding calls fail outside a handler | [Environment access and global-scope restrictions](https://developers.cloudflare.com/workers/runtime-apis/bindings/#how-to-access-env) |
| A service target is unavailable or incorrect | [Service bindings](https://developers.cloudflare.com/workers/runtime-apis/bindings/service-bindings/) and [multi-Worker development](https://developers.cloudflare.com/workers/local-development/multi-workers/) |
| Local data, secrets, or available bindings differ from deployment | [Development mode support](https://developers.cloudflare.com/workers/local-development/bindings-per-env/), [local data](https://developers.cloudflare.com/workers/local-development/local-data/), and [local secrets](https://developers.cloudflare.com/workers/local-development/environment-variables/) |
| KV reads appear stale or return no value | [How KV works](https://developers.cloudflare.com/kv/concepts/how-kv-works/); verify the namespace and handle missing values |
| Resource-specific API errors or limits | Follow the product from the [binding catalog](https://developers.cloudflare.com/workers/runtime-apis/bindings/) to its troubleshooting and limits documentation |
| Worker resource limits or unexpected charges | [Workers limits](https://developers.cloudflare.com/workers/platform/limits/) and [pricing](https://developers.cloudflare.com/workers/platform/pricing/); check each bound product's limits and pricing separately |
| Need to inspect configuration, resources, deployments, or logs | [Wrangler commands](https://developers.cloudflare.com/workers/wrangler/commands/) and [real-time logs](https://developers.cloudflare.com/workers/observability/logs/real-time-logs/) |

Never log secret values or return the environment object in a response. Inspect names and configuration without exposing credentials. Check the selected resource and development mode before issuing debugging commands that could mutate remote data.

# Binding Patterns

Choose the interaction and lifecycle first, then retrieve the implementation guide.

| Task | Current documentation |
| --- | --- |
| Choose HTTP forwarding or RPC between Workers | [Service bindings](https://developers.cloudflare.com/workers/runtime-apis/bindings/service-bindings/) |
| Forward Requests and Responses through a service binding | [Service bindings over HTTP](https://developers.cloudflare.com/workers/runtime-apis/bindings/service-bindings/http/) |
| Expose callable methods with `WorkerEntrypoint` | [Service bindings over RPC](https://developers.cloudflare.com/workers/runtime-apis/bindings/service-bindings/rpc/) and [RPC TypeScript](https://developers.cloudflare.com/workers/runtime-apis/rpc/typescript/) |
| Run connected Workers during development | [Developing with multiple Workers](https://developers.cloudflare.com/workers/local-development/multi-workers/) |
| Test handlers against configured bindings and mock dependencies | [Workers Vitest configuration](https://developers.cloudflare.com/workers/testing/vitest-integration/configuration/) and [test APIs](https://developers.cloudflare.com/workers/testing/vitest-integration/test-apis/) |
| Select KV, D1, R2, or Durable Objects | [Storage options](https://developers.cloudflare.com/workers/platform/storage-options/) |
| Keep clients current when bindings change | [Binding lifecycle](https://developers.cloudflare.com/workers/runtime-apis/bindings/#making-changes-to-bindings) |
| Manage credentials used by external API clients | [Secrets](https://developers.cloudflare.com/workers/configuration/secrets/) |

Use service bindings for internal Worker calls when appropriate, and choose HTTP or RPC based on the interface being exposed. A service binding does not replace application-level authorization for the caller's requested operation.

Choose storage based on access patterns and consistency requirements, not copied size or latency thresholds. Parallelize independent binding operations when useful; preserve ordering where one operation depends on another's result.

Avoid retaining clients derived from mutable bindings across requests without accounting for binding updates. Importing `env` is supported, but binding I/O still requires an appropriate execution context; follow the lifecycle guide rather than assuming all global access is forbidden.

# Cloudflare Workers Bindings

Bindings grant a Worker access to configured resources through its environment. Prefer a product's binding for supported operations inside Workers; use the REST API when the caller or operation requires it.

Read the relevant current documentation before implementing. These references route to maintained APIs and configuration rather than copying binding catalogs, type tables, or limits.

## Start here

- [Bindings overview and catalog](https://developers.cloudflare.com/workers/runtime-apis/bindings/): capability model, available products, environment access, and binding lifecycle.
- [Storage options](https://developers.cloudflare.com/workers/platform/storage-options/): choose storage from consistency, query, and coordination requirements.
- [api.md](./api.md): environment access, generated types, and product APIs.
- [configuration.md](./configuration.md): binding configuration, environments, secrets, and local development.
- [patterns.md](./patterns.md): Worker-to-Worker calls, testing, and resource selection.
- [gotchas.md](./gotchas.md): missing bindings, stale clients, development differences, and limits.

Treat each binding as a capability granted to code. Select only the resources the Worker needs, and confirm which environment and resource each binding targets before using it.

# Bot Management API

## Workers: BotManagement Interface

```typescript
interface BotManagement {
  score: number;              // 1-99 (Enterprise), 0 if not computed
  verifiedBot: boolean;       // Is verified bot
  staticResource: boolean;    // Serves static resource
  ja3Hash: string;            // JA3 fingerprint (Enterprise, HTTPS only)
  ja4: string;                // JA4 fingerprint (Enterprise, HTTPS only)
  jsDetection?: {
    passed: boolean;          // Passed JS detection (if enabled)
  };
  detectionIds: number[];     // Heuristic detection IDs
  corporateProxy?: boolean;   // From corporate proxy (Enterprise)
}

// DEPRECATED: Use botManagement.score instead
// request.cf.clientTrustScore (legacy, duplicate of botManagement.score)

// Access via request.cf
import type { IncomingRequestCfProperties } from '@cloudflare/workers-types';

export default {
  async fetch(request: Request): Promise<Response> {
    const cf = request.cf as IncomingRequestCfProperties | undefined;
    const botMgmt = cf?.botManagement;
    
    if (!botMgmt) return fetch(request);
    if (botMgmt.verifiedBot) return fetch(request); // Allow verified bots
    if (botMgmt.score === 1) return new Response('Blocked', { status: 403 });
    if (botMgmt.score < 30) return new Response('Challenge required', { status: 429 });
    
    return fetch(request);
  }
};
```

## WAF Fields Reference

```txt
# Score fields
cf.bot_management.score                    # 0-99 (0 = not computed)
cf.bot_management.verified_bot             # boolean
cf.bot_management.static_resource          # boolean
cf.bot_management.ja3_hash                 # string (Enterprise)
cf.bot_management.ja4                      # string (Enterprise)
cf.bot_management.detection_ids            # array
cf.bot_management.js_detection.passed      # boolean
cf.bot_management.corporate_proxy          # boolean (Enterprise)
cf.verified_bot_category                   # string

# Workers equivalent
request.cf.botManagement.score
request.cf.botManagement.verifiedBot
request.cf.botManagement.ja3Hash
request.cf.botManagement.ja4
request.cf.botManagement.jsDetection.passed
request.cf.verifiedBotCategory
```

## JA4 Signals (Enterprise)

```typescript
import type { IncomingRequestCfProperties } from '@cloudflare/workers-types';

interface JA4Signals {
  // Ratios (0.0-1.0)
  heuristic_ratio_1h?: number;  // Fraction flagged by heuristics
  browser_ratio_1h?: number;    // Fraction from real browsers  
  cache_ratio_1h?: number;      // Fraction hitting cache
  h2h3_ratio_1h?: number;       // Fraction using HTTP/2 or HTTP/3
  // Ranks (relative position in distribution)
  uas_rank_1h?: number;         // User-Agent diversity rank
  paths_rank_1h?: number;       // Path diversity rank
  reqs_rank_1h?: number;        // Request volume rank
  ips_rank_1h?: number;         // IP diversity rank
  // Quantiles (0.0-1.0, percentile in distribution)
  reqs_quantile_1h?: number;    // Request volume quantile
  ips_quantile_1h?: number;     // IP count quantile
}

export default {
  async fetch(request: Request): Promise<Response> {
    const cf = request.cf as IncomingRequestCfProperties | undefined;
    const ja4Signals = cf?.ja4Signals as JA4Signals | undefined;
    
    if (!ja4Signals) return fetch(request); // Not available for HTTP or Worker routing
    
    // Check for anomalous behavior
    // High heuristic_ratio or low browser_ratio = suspicious
    const heuristicRatio = ja4Signals.heuristic_ratio_1h ?? 0;
    const browserRatio = ja4Signals.browser_ratio_1h ?? 0;
    
    if (heuristicRatio > 0.5 || browserRatio < 0.3) {
      return new Response('Suspicious traffic', { status: 403 });
    }
    
    return fetch(request);
  }
};
```

## Common Patterns

See [patterns.md](./patterns.md) for Workers examples: mobile app allowlisting, corporate proxy exemption, datacenter detection, conditional delay, and more.

## Bot Analytics

### Access Locations
- Dashboard: Security > Bots (old) or Security > Analytics > Bot analysis (new)
- GraphQL API for programmatic access
- Security Events & Security Analytics
- Logpush/Logpull

### Available Data
- **Enterprise BM**: Bot scores (1-99), bot score source, distribution
- **Pro/Business**: Bot groupings (automated, likely automated, likely human)
- Top attributes: IPs, paths, user agents, countries
- Detection sources: Heuristics, ML, AD, JSD
- Verified bot categories

### Time Ranges
- **Enterprise BM**: Up to 1 week at a time, 30 days history
- **Pro/Business**: Up to 72 hours at a time, 30 days history
- Real-time in most cases, adaptive sampling (1-10% depending on volume)

## Logpush Fields

```txt
BotScore              # 1-99 or 0 if not computed
BotScoreSrc           # Detection engine (ML, Heuristics, etc.)
BotTags               # Classification tags
BotDetectionIDs       # Heuristic detection IDs
```

**BotScoreSrc values:**
- `"Heuristics"` - Known fingerprint
- `"Machine Learning"` - ML model
- `"Anomaly Detection"` - Baseline anomaly
- `"JS Detection"` - JavaScript check
- `"Cloudflare Service"` - Zero Trust
- `"Not Computed"` - Score = 0

Access via Logpush (stream to cloud storage/SIEM), Logpull (API to fetch logs), or GraphQL API (query analytics data).

## Testing with Miniflare

Miniflare provides mock botManagement data for local development:

**Default values:**
- `score: 99` (human)
- `verifiedBot: false`
- `corporateProxy: false`
- `ja3Hash: "25b4882c2bcb50cd6b469ff28c596742"`
- `staticResource: false`
- `detectionIds: []`

**Override in tests:**
```typescript
import { getPlatformProxy } from 'wrangler';

const { cf, dispose } = await getPlatformProxy();
// cf.botManagement is frozen mock object
expect(cf.botManagement.score).toBe(99);
```

For custom test data, mock request.cf in your test setup.

# Bot Management Configuration

## Product Tiers

**Note:** Dashboard paths differ between old and new UI:
- **New:** Security > Settings > Filter "Bot traffic"
- **Old:** Security > Bots

Both UIs access same settings.

### Bot Score Groupings (Pro/Business)

Pro/Business users see bot score groupings instead of granular 1-99 scores:

| Score | Grouping | Meaning |
|-------|----------|---------|
| 0 | Not computed | Bot Management didn't run |
| 1 | Automated | Definite bot (heuristic match) |
| 2-29 | Likely automated | Probably bot (ML detection) |
| 30-99 | Likely human | Probably human |
| N/A | Verified bot | Allowlisted good bot |

Enterprise plans get granular 1-99 scores for custom thresholds.

### Bot Fight Mode (Free)
- Auto-blocks definite bots (score=1), excludes verified bots by default
- JavaScript Detections always enabled, no configuration options

### Super Bot Fight Mode (Pro/Business)
```txt
Dashboard: Security > Bots > Configure
- Definitely automated: Block/Challenge
- Likely automated: Challenge/Allow  
- Verified bots: Allow (recommended)
- Static resource protection: ON (may block mail clients)
- JavaScript Detections: Optional
```

### Bot Management for Enterprise
```txt
Dashboard: Security > Bots > Configure > Auto-updates: ON (recommended)

# Template 1: Block definite bots
(cf.bot_management.score eq 1 and not cf.bot_management.verified_bot and not cf.bot_management.static_resource)
Action: Block

# Template 2: Challenge likely bots
(cf.bot_management.score ge 2 and cf.bot_management.score le 29 and not cf.bot_management.verified_bot and not cf.bot_management.static_resource)
Action: Managed Challenge
```

## JavaScript Detections Setup

### Enable via Dashboard
```txt
Security > Bots > Configure Bot Management > JS Detections: ON

Update CSP: script-src 'self' /cdn-cgi/challenge-platform/;
```

### Manual JS Injection (API)
```html
<script>
function jsdOnload() {
  window.cloudflare.jsd.executeOnce({ callback: function(result) { console.log('JSD:', result); } });
}
</script>
<script src="/cdn-cgi/challenge-platform/scripts/jsd/api.js?onload=jsdOnload" async></script>
```

**Use API for**: Selective deployment on specific pages  
**Don't combine**: Zone-wide toggle + manual injection

### WAF Rules for JSD
```txt
# NEVER use on first page visit (needs HTML page first)
(not cf.bot_management.js_detection.passed and http.request.uri.path eq "/api/user/create" and http.request.method eq "POST" and not cf.bot_management.verified_bot)
Action: Managed Challenge (always use Managed Challenge, not Block)
```

### Limitations
- First request won't have JSD data (needs HTML page first)
- Strips ETags from HTML responses
- Not supported with CSP via `<meta>` tags
- Websocket endpoints not supported
- Native mobile apps won't pass
- cf_clearance cookie: 15-minute lifespan, max 4096 bytes

## __cf_bm Cookie

Cloudflare sets `__cf_bm` cookie to smooth bot scores across user sessions:

- **Purpose:** Reduces false positives from score volatility
- **Scope:** Per-domain, HTTP-only
- **Lifespan:** Session duration
- **Privacy:** No PII—only session classification
- **Automatic:** No configuration required

Bot scores for repeat visitors consider session history via this cookie.

## Static Resource Protection

**File Extensions**: ico, jpg, png, jpeg, gif, css, js, tif, tiff, bmp, pict, webp, svg, svgz, class, jar, txt, csv, doc, docx, xls, xlsx, pdf, ps, pls, ppt, pptx, ttf, otf, woff, woff2, eot, eps, ejs, swf, torrent, midi, mid, m3u8, m4a, mp3, ogg, ts  
**Plus**: `/.well-known/` path (all files)

```txt
# Exclude static resources from bot rules
(cf.bot_management.score lt 30 and not cf.bot_management.static_resource)
```

**WARNING**: May block mail clients fetching static images

## JA3/JA4 Fingerprinting (Enterprise)

```txt
# Block specific attack fingerprint
(cf.bot_management.ja3_hash eq "8b8e3d5e3e8b3d5e")

# Allow mobile app by fingerprint
(cf.bot_management.ja4 eq "your_mobile_app_fingerprint")
```

Only available for HTTPS/TLS traffic. Missing for Worker-routed traffic or HTTP requests.

## Verified Bot Categories

```txt
# Allow search engines only
(cf.verified_bot_category eq "Search Engine Crawler")

# Block AI crawlers
(cf.verified_bot_category eq "AI Crawler")
Action: Block

# Or use dashboard: Security > Settings > Bot Management > Block AI Bots
```

| Category | String Value | Example |
|----------|--------------|---------|
| AI Crawler | `AI Crawler` | GPTBot, Claude-Web |
| AI Assistant | `AI Assistant` | Perplexity-User, DuckAssistBot |
| AI Search | `AI Search` | OAI-SearchBot |
| Accessibility | `Accessibility` | Accessible Web Bot |
| Academic Research | `Academic Research` | Library of Congress |
| Advertising & Marketing | `Advertising & Marketing` | Google Adsbot |
| Aggregator | `Aggregator` | Pinterest, Indeed |
| Archiver | `Archiver` | Internet Archive, CommonCrawl |
| Feed Fetcher | `Feed Fetcher` | RSS/Podcast updaters |
| Monitoring & Analytics | `Monitoring & Analytics` | Uptime monitors |
| Page Preview | `Page Preview` | Facebook/Slack link preview |
| SEO | `Search Engine Optimization` | Google Lighthouse |
| Security | `Security` | Vulnerability scanners |
| Social Media Marketing | `Social Media Marketing` | Brandwatch |
| Webhooks | `Webhooks` | Payment processors |
| Other | `Other` | Uncategorized bots |

## Best Practices

- **ML Auto-Updates**: Enable on Enterprise for latest models
- **Start with Managed Challenge**: Test before blocking
- **Always exclude verified bots**: Use `not cf.bot_management.verified_bot`
- **Exempt corporate proxies**: For B2B traffic via `cf.bot_management.corporate_proxy`
- **Use static resource exception**: Improves performance, reduces overhead

# Bot Management Gotchas

## Common Errors

### "Bot Score = 0"

**Cause:** Bot Management didn't run (internal Cloudflare request, Worker routing to zone (Orange-to-Orange), or request handled before BM (Redirect Rules, etc.))  
**Solution:** Check request flow and ensure Bot Management runs in request lifecycle

### "JavaScript Detections Not Working"

**Cause:** `js_detection.passed` always false or undefined due to: CSP headers don't allow `/cdn-cgi/challenge-platform/`, using on first page visit (needs HTML page first), ad blockers or disabled JS, JSD not enabled in dashboard, or using Block action (must use Managed Challenge)  
**Solution:** Add CSP header `Content-Security-Policy: script-src 'self' /cdn-cgi/challenge-platform/;` and ensure JSD is enabled with Managed Challenge action

### "False Positives (Legitimate Users Blocked)"

**Cause:** Bot detection incorrectly flagging legitimate users  
**Solution:** Check Bot Analytics for affected IPs/paths, identify detection source (ML, Heuristics, etc.), create exception rule like `(cf.bot_management.score lt 30 and http.request.uri.path eq "/problematic-path")` with Action: Skip (Bot Management), or allowlist by IP/ASN/country

### "False Negatives (Bots Not Caught)"

**Cause:** Bots bypassing detection  
**Solution:** Lower score threshold (30 → 50), enable JavaScript Detections, add JA3/JA4 fingerprinting rules, or use rate limiting as fallback

### "Verified Bot Blocked"

**Cause:** Search engine bot blocked by WAF Managed Rules (not just Bot Management)  
**Solution:** Create WAF exception for specific rule ID and verify bot via reverse DNS

### "Yandex Bot Blocked During IP Update"

**Cause:** Yandex updates bot IPs; new IPs unrecognized for 48h during propagation  
**Solution:** 
1. Check Security Events for specific WAF rule ID blocking Yandex
2. Create WAF exception:
   ```txt
   (http.user_agent contains "YandexBot" and ip.src in {<yandex-ip-range>})
   Action: Skip (WAF Managed Ruleset)
   ```
3. Monitor Bot Analytics for 48h
4. Remove exception after propagation completes

Issue resolves automatically after 48h. Contact Cloudflare Support if persists.

### "JA3/JA4 Missing"

**Cause:** Non-HTTPS traffic, Worker routing traffic, Orange-to-Orange traffic via Worker, or Bot Management skipped  
**Solution:** JA3/JA4 only available for HTTPS/TLS traffic; check request routing

**JA3/JA4 Not User-Unique:** Same browser/library version = same fingerprint
- Don't use for user identification
- Use for client profiling only
- Fingerprints change with browser updates

## Bot Verification Methods

Cloudflare verifies bots via:

1. **Reverse DNS (IP validation):** Traditional method—bot IP resolves to expected domain
2. **Web Bot Auth:** Modern cryptographic verification—faster propagation

When `verifiedBot=true`, bot passed at least one method.

**Inactive verified bots:** IPs removed after 24h of no traffic.

## Detection Engine Behavior

| Engine | Score | Timing | Plan | Notes |
|--------|-------|--------|------|-------|
| Heuristics | Always 1 | Immediate | All | Known fingerprints—overrides ML |
| ML | 1-99 | Immediate | All | Majority of detections |
| Anomaly Detection | Influences | After baseline | Enterprise | Optional, baseline analysis |
| JavaScript Detections | Pass/fail | After JS | Pro+ | Headless browser detection |
| Cloudflare Service | N/A | N/A | Enterprise | Zero Trust internal source |

**Priority:** Heuristics > ML—if heuristic matches, score=1 regardless of ML.

## Limits

| Limit | Value | Notes |
|-------|-------|-------|
| Bot Score = 0 | Means not computed | Not score = 100 |
| First request JSD data | May not be available | JSD data appears on subsequent requests |
| Score accuracy | Not 100% guaranteed | False positives/negatives possible |
| JSD on first HTML page visit | Not supported | Requires subsequent page load |
| JSD requirements | JavaScript-enabled browser | Won't work with JS disabled or ad blockers |
| JSD ETag stripping | Strips ETags from HTML responses | May affect caching behavior |
| JSD CSP compatibility | Requires specific CSP | Not compatible with some CSP configurations |
| JSD meta CSP tags | Not supported | Must use HTTP headers |
| JSD WebSocket support | Not supported | WebSocket endpoints won't work with JSD |
| JSD mobile app support | Native apps won't pass | Only works in browsers |
| JA3/JA4 traffic type | HTTPS/TLS only | Not available for non-HTTPS traffic |
| JA3/JA4 Worker routing | Missing for Worker-routed traffic | Check request routing |
| JA3/JA4 uniqueness | Not unique per user | Shared by clients with same browser/library |
| JA3/JA4 stability | Can change with updates | Browser/library updates affect fingerprints |
| WAF custom rules (Free) | 5 | Varies by plan |
| WAF custom rules (Pro) | 20 | Varies by plan |
| WAF custom rules (Business) | 100 | Varies by plan |
| WAF custom rules (Enterprise) | 1,000+ | Varies by plan |
| Workers CPU time | Varies by plan | Applies to bot logic |
| Bot Analytics sampling | 1-10% adaptive | High-volume zones sampled more aggressively |
| Bot Analytics history | 30 days max | Historical data retention limit |
| CSP requirements for JSD | Must allow `/cdn-cgi/challenge-platform/` | Required for JSD to function |

### Plan Restrictions

| Feature | Free | Pro/Business | Enterprise |
|---------|------|--------------|------------|
| Granular scores (1-99) | No | No | Yes |
| JA3/JA4 | No | No | Yes |
| Anomaly Detection | No | No | Yes |
| Corporate Proxy detection | No | No | Yes |
| Verified bot categories | Limited | Limited | Full |
| Custom WAF rules | 5 | 20/100 | 1,000+ |

# Bot Management Patterns

## E-commerce Protection

```txt
# High security for checkout
(cf.bot_management.score lt 50 and http.request.uri.path in {"/checkout" "/cart/add"} and not cf.bot_management.verified_bot and not cf.bot_management.corporate_proxy)
Action: Managed Challenge
```

## API Protection

```txt
# Protect API with JS detection + score
(http.request.uri.path matches "^/api/" and (cf.bot_management.score lt 30 or not cf.bot_management.js_detection.passed) and not cf.bot_management.verified_bot)
Action: Block
```

## SEO-Friendly Bot Handling

```txt
# Allow search engine crawlers
(cf.bot_management.score lt 30 and not cf.verified_bot_category in {"Search Engine Crawler"})
Action: Managed Challenge
```

## Block AI Scrapers

```txt
# Block training crawlers only (allow AI assistants/search)
(cf.verified_bot_category eq "AI Crawler")
Action: Block

# Block all AI-related bots (training + assistants + search)
(cf.verified_bot_category in {"AI Crawler" "AI Assistant" "AI Search"})
Action: Block

# Allow AI Search, block AI Crawler and AI Assistant
(cf.verified_bot_category in {"AI Crawler" "AI Assistant"})
Action: Block

# Or use dashboard: Security > Settings > Bot Management > Block AI Bots
```

## Rate Limiting by Bot Score

```txt
# Stricter limits for suspicious traffic
(cf.bot_management.score lt 50)
Rate: 10 requests per 10 seconds

(cf.bot_management.score ge 50)
Rate: 100 requests per 10 seconds
```

## Mobile App Allowlisting

```txt
# Identify mobile app by JA3/JA4
(cf.bot_management.ja4 in {"fingerprint1" "fingerprint2"})
Action: Skip (all remaining rules)
```

## Datacenter Detection

```typescript
import type { IncomingRequestCfProperties } from '@cloudflare/workers-types';

// Low score + not corporate proxy = likely datacenter bot
export default {
  async fetch(request: Request): Promise<Response> {
    const cf = request.cf as IncomingRequestCfProperties | undefined;
    const botMgmt = cf?.botManagement;
    
    if (botMgmt?.score && botMgmt.score < 30 && 
        !botMgmt.corporateProxy && !botMgmt.verifiedBot) {
      return new Response('Datacenter traffic blocked', { status: 403 });
    }
    
    return fetch(request);
  }
};
```

## Conditional Delay (Tarpit)

```typescript
import type { IncomingRequestCfProperties } from '@cloudflare/workers-types';

// Add delay proportional to bot suspicion
export default {
  async fetch(request: Request): Promise<Response> {
    const cf = request.cf as IncomingRequestCfProperties | undefined;
    const botMgmt = cf?.botManagement;
    
    if (botMgmt?.score && botMgmt.score < 50 && !botMgmt.verifiedBot) {
      // Delay: 0-2 seconds for scores 50-0
      const delayMs = Math.max(0, (50 - botMgmt.score) * 40);
      await new Promise(r => setTimeout(r, delayMs));
    }
    
    return fetch(request);
  }
};
```

## Layered Defense

```txt
1. Bot Management (score-based)
2. JavaScript Detections (for JS-capable clients)
3. Rate Limiting (fallback protection)
4. WAF Managed Rules (OWASP, etc.)
```

## Progressive Enhancement

```txt
Public content: High threshold (score < 10)
Authenticated: Medium threshold (score < 30)
Sensitive: Low threshold (score < 50) + JSD
```

## Zero Trust for Bots

```txt
1. Default deny (all scores < 30)
2. Allowlist verified bots
3. Allowlist mobile apps (JA3/JA4)
4. Allowlist corporate proxies
5. Allowlist static resources
```

## Workers: Score + JS Detection

```typescript
import type { IncomingRequestCfProperties } from '@cloudflare/workers-types';

export default {
  async fetch(request: Request): Promise<Response> {
    const cf = request.cf as IncomingRequestCfProperties | undefined;
    const botMgmt = cf?.botManagement;
    const url = new URL(request.url);
    
    if (botMgmt?.staticResource) return fetch(request); // Skip static
    
    // API endpoints: require JS detection + good score
    if (url.pathname.startsWith('/api/')) {
      const jsDetectionPassed = botMgmt?.jsDetection?.passed ?? false;
      const score = botMgmt?.score ?? 100;
      
      if (!jsDetectionPassed || score < 30) {
        return new Response('Unauthorized', { status: 401 });
      }
    }
    
    return fetch(request);
  }
};
```

## Rate Limiting by JWT Claim + Bot Score

```txt
# Enterprise: Combine bot score with JWT validation
Rate limiting > Custom rules
- Field: lookup_json_string(http.request.jwt.claims["{config_id}"][0], "sub")
- Matches: user ID claim
- Additional condition: cf.bot_management.score lt 50
```

## WAF Integration Points

- **WAF Custom Rules**: Primary enforcement mechanism
- **Rate Limiting Rules**: Bot score as dimension, stricter limits for low scores
- **Transform Rules**: Pass score to origin via custom header
- **Workers**: Programmatic bot logic, custom scoring algorithms
- **Page Rules / Configuration Rules**: Zone-level overrides, path-specific settings

## See Also

- [gotchas.md](./gotchas.md) - Common errors, false positives/negatives, limitations

# Cloudflare Bot Management

Enterprise-grade bot detection, protection, and mitigation using ML/heuristics, bot scores, JavaScript detections, and verified bot handling.

## Overview

Bot Management provides multi-tier protection:
- **Free (Bot Fight Mode)**: Auto-blocks definite bots, no config
- **Pro/Business (Super Bot Fight Mode)**: Configurable actions, static resource protection, analytics groupings
- **Enterprise (Bot Management)**: Granular 1-99 scores, WAF integration, JA3/JA4 fingerprinting, Workers API, Advanced Analytics

## Quick Start

```txt
# Dashboard: Security > Bots
# Enterprise: Deploy rule template
(cf.bot_management.score eq 1 and not cf.bot_management.verified_bot) → Block
(cf.bot_management.score le 29 and not cf.bot_management.verified_bot) → Managed Challenge
```

## What Do You Need?

```txt
├─ Initial setup → configuration.md
│   ├─ Free tier → "Bot Fight Mode"
│   ├─ Pro/Business → "Super Bot Fight Mode"
│   └─ Enterprise → "Bot Management for Enterprise"
├─ Workers API integration → api.md
├─ WAF rules → patterns.md
├─ Debugging → gotchas.md
└─ Analytics → api.md#bot-analytics
```

## Reading Order

| Task | Files to Read |
|------|---------------|
| Enable bot protection | README → configuration.md |
| Workers bot detection | README → api.md |
| WAF rule templates | README → patterns.md |
| Debug bot issues | gotchas.md |
| Advanced analytics | api.md#bot-analytics |

## Core Concepts

**Bot Scores**: 1-99 (1 = definitely automated, 99 = definitely human). Threshold: <30 indicates bot traffic. Enterprise gets granular 1-99; Pro/Business get groupings only.

**Detection Engines**: Heuristics (known fingerprints, assigns score=1), ML (majority of detections, supervised learning on billions of requests), Anomaly Detection (optional, baseline traffic analysis), JavaScript Detections (headless browser detection).

**Verified Bots**: Allowlisted good bots (search engines, AI crawlers) verified via reverse DNS or Web Bot Auth. Access via `cf.bot_management.verified_bot` or `cf.verified_bot_category`.

## Platform Limits

| Plan | Bot Scores | JA3/JA4 | Custom Rules | Analytics Retention |
|------|------------|---------|--------------|---------------------|
| Free | No (auto-block only) | No | 5 | N/A (no analytics) |
| Pro/Business | Groupings only | No | 20/100 | 30 days (72h at a time) |
| Enterprise | 1-99 granular | Yes | 1,000+ | 30 days (1 week at a time) |

## Basic Patterns

```typescript
// Workers: Check bot score
export default {
  async fetch(request: Request): Promise<Response> {
    const botScore = request.cf?.botManagement?.score;
    if (botScore && botScore < 30 && !request.cf?.botManagement?.verifiedBot) {
      return new Response('Bot detected', { status: 403 });
    }
    return fetch(request);
  }
};
```

```txt
# WAF: Block definite bots
(cf.bot_management.score eq 1 and not cf.bot_management.verified_bot)

# WAF: Protect sensitive endpoints
(cf.bot_management.score lt 50 and http.request.uri.path in {"/login" "/checkout"} and not cf.bot_management.verified_bot)
```

## In This Reference

- [configuration.md](./configuration.md) - Product tiers, WAF rule setup, JavaScript Detections, ML auto-updates
- [api.md](./api.md) - Workers BotManagement interface, WAF fields, JA4 Signals
- [patterns.md](./patterns.md) - E-commerce, API protection, mobile app allowlisting, SEO-friendly handling
- [gotchas.md](./gotchas.md) - False positives/negatives, score=0 issues, JSD limitations, CSP requirements

## See Also

- [waf](../waf/) - WAF custom rules for bot enforcement
- [workers](https://developers.cloudflare.com/workers/) - Workers request.cf.botManagement API
- [api-shield](../api-shield/) - API-specific bot protection

# Browser Run APIs

Read the guide for the chosen interface for request schemas, return types, authentication, and supported options. Keep Quick Actions and browser session APIs distinct when adapting examples.

| Task | Documentation |
|------|---------------|
| Screenshots, PDFs, HTML, scraping, or structured extraction | [Quick Actions](https://developers.cloudflare.com/browser-run/quick-actions/) — links to each action's request options and examples for REST or Workers bindings |
| Automate a browser in Workers with Puppeteer | [Puppeteer](https://developers.cloudflare.com/browser-run/puppeteer/) — Cloudflare package, browser operations, and session APIs |
| Automate a browser in Workers with Playwright | [Playwright](https://developers.cloudflare.com/browser-run/playwright/) — Cloudflare package, locators, storage state, and tracing |
| Control a remote browser from an external runtime | [CDP](https://developers.cloudflare.com/browser-run/cdp/) — session endpoints and links to Puppeteer, Playwright, and other clients |

The product rename does not imply a rename of API paths or token permissions. Use the identifiers shown in the selected guide.

# Browser Run Configuration

Check the project's runtime, installed client and Wrangler versions, and compatibility date before adapting setup instructions. Cloudflare's packages for Workers and standard clients connecting over CDP have different setup requirements.

| Task | Documentation |
|------|---------------|
| Start a project or configure REST authentication | [Get started](https://developers.cloudflare.com/browser-run/get-started/) — Quick Actions and browser session setup |
| Configure a Worker or choose a development mode | [Wrangler reference](https://developers.cloudflare.com/browser-run/reference/wrangler/) — browser bindings, compatibility requirements, and local/remote development |
| Install or update a Workers browser client | [Puppeteer](https://developers.cloudflare.com/browser-run/puppeteer/) or [Playwright](https://developers.cloudflare.com/browser-run/playwright/) — package-specific setup and supported versions |
| Connect from a script, server, or CI outside Workers | [CDP](https://developers.cloudflare.com/browser-run/cdp/) — authentication and client integration guides |

Development support depends on the selected interface. Follow its current guidance rather than applying one remote-mode requirement to all Browser Run workflows.

# Browser Run Troubleshooting

Identify the integration and observed failure before changing timeouts or concurrency. A request-rate limit, exhausted browser time, and a closed session require different responses.

| Concern | Documentation |
|---------|---------------|
| Quotas, launch rates, concurrency, and session timeouts | [Limits](https://developers.cloudflare.com/browser-run/limits/) — check the current plan and integration-specific limits |
| Browser hours and concurrent-browser charges | [Pricing](https://developers.cloudflare.com/browser-run/pricing/) — distinguish Quick Actions from browser sessions |
| Missing bindings, action failures, or unsupported behavior | [FAQ](https://developers.cloudflare.com/browser-run/faq/) — diagnose the reported error and runtime constraints |
| Puppeteer page evaluation cannot access outer variables | [JavaScript execution](https://pptr.dev/guides/javascript-execution) — browser execution context, passing arguments, and returned values |
| Block resources or handle intercepted Puppeteer requests | [Request interception](https://pptr.dev/guides/network-interception) — continue, respond, or abort requests and avoid duplicate handling |
| Unexpected disconnects or session loss | [Browser close reasons](https://developers.cloudflare.com/browser-run/reference/browser-close-reasons/) — inspect the recorded close reason before choosing recovery |
| Development or compatibility failures | [Wrangler reference](https://developers.cloudflare.com/browser-run/reference/wrangler/) — verify binding configuration and interface-specific development support |

Before increasing concurrency, check session cleanup and whether the workload can reuse browsers with appropriate isolation; see [patterns.md](patterns.md). Retrieve current limits and pricing when sizing a workload rather than relying on fixed tier tables.

# Browser Run Patterns

Use the current examples for the selected integration instead of translating between Puppeteer, Playwright, and Quick Actions by changing method names.

| Task | Documentation |
|------|---------------|
| Implement screenshots, PDFs, or extraction | [Quick Actions](https://developers.cloudflare.com/browser-run/quick-actions/) — choose the action and follow its example |
| Build custom interactions | [Puppeteer](https://developers.cloudflare.com/browser-run/puppeteer/) or [Playwright](https://developers.cloudflare.com/browser-run/playwright/) — browser automation examples |
| Reconnect across requests | [Reuse sessions](https://developers.cloudflare.com/browser-run/features/reuse-sessions/) — disconnect/reconnect lifecycle and when to use Durable Objects for stateful ownership |
| Share browser capacity while isolating users | [Concurrency and session isolation](https://developers.cloudflare.com/browser-run/limits/#how-can-i-manage-concurrency-and-session-isolation-with-browser-run) — tabs, browser contexts, and capacity tradeoffs |

Quick Actions manage their own session lifecycle. For sessions managed by the application, close pages and browsers on completion or failure. If reuse is intentional, follow the client's disconnect/reconnect semantics and handle expired sessions; closing the browser ends it. Keep cookies and storage isolated between users, and coordinate ownership when several requests can reconnect to the same session.

# Browser Run (formerly Browser Rendering)

Use Browser Run for screenshots, PDFs, rendered content extraction, and browser automation. Read the relevant current documentation before implementing; use the [documentation index](https://developers.cloudflare.com/browser-run/llms.txt) to discover additional guides.

Choose the integration by the work and runtime:

- For a self-contained screenshot, PDF, or extraction, start with Quick Actions. They are available through REST and Workers bindings; check the chosen action's supported interface.
- For multi-step interactions or persistent state, use browser sessions. In Workers, use Cloudflare's Puppeteer or Playwright package; from external scripts or CI, use the CDP integration.
- When adapting existing automation, preserve its library where supported and check installed versions against the corresponding guide.

Read only the reference needed for the task:

| Task | Reference |
|------|-----------|
| Set up bindings, dependencies, or development | [configuration.md](configuration.md) |
| Select an endpoint or browser client API | [api.md](api.md) |
| Implement a workflow or manage reusable sessions | [patterns.md](patterns.md) |
| Diagnose failures or plan capacity and cost | [gotchas.md](gotchas.md) |

# C3 CLI Reference

## Invocation

```bash
npm create cloudflare@latest [name] [-- flags]  # NPM requires --
yarn create cloudflare [name] [flags]
pnpm create cloudflare@latest [name] [-- flags]
```

## Core Flags

| Flag | Values | Description |
|------|--------|-------------|
| `--type` | `hello-world`, `web-app`, `demo`, `pre-existing`, `remote-template` | Application type |
| `--platform` | `workers` (default), `pages` | Target platform |
| `--framework` | `next`, `remix`, `astro`, `react-router`, `solid`, `svelte`, `qwik`, `vue`, `angular`, `hono` | Web framework (requires `--type=web-app`) |
| `--lang` | `ts`, `js`, `python` | Language (for `--type=hello-world`) |
| `--ts` / `--no-ts` | - | TypeScript for web apps |

## Deployment Flags

| Flag | Description |
|------|-------------|
| `--deploy` / `--no-deploy` | Deploy immediately (prompts interactive, skips in CI) |
| `--git` / `--no-git` | Initialize git (default: yes) |
| `--open` | Open browser after deploy |

## Advanced Flags

| Flag | Description |
|------|-------------|
| `--template=user/repo` | GitHub template or local path |
| `--existing-script=./src/worker.ts` | Existing script (requires `--type=pre-existing`) |
| `--category=ai\|database\|realtime` | Demo filter (requires `--type=demo`) |
| `--experimental` | Enable experimental features |
| `--wrangler-defaults` | Skip wrangler prompts |

## Environment Variables

```bash
CLOUDFLARE_API_TOKEN=xxx    # For deployment
CLOUDFLARE_ACCOUNT_ID=xxx   # Account ID
CF_TELEMETRY_DISABLED=1     # Disable telemetry
```

## Exit Codes

`0` success, `1` user abort, `2` error

## Examples

For framework apps, follow [Framework Setup](README.md#framework-setup).

```bash
# TypeScript Worker
npm create cloudflare@latest my-api -- --type=hello-world --lang=ts --no-deploy

# Astro blog
npm create cloudflare@latest my-blog -- --type=web-app --framework=astro --ts --deploy

# CI: non-interactive
npm create cloudflare@latest my-api -- --type=hello-world --lang=ts --no-git --no-deploy

# GitHub template
npm create cloudflare@latest -- --template=cloudflare/templates/worker-openapi

# Convert existing project
npm create cloudflare@latest . -- --type=pre-existing --existing-script=./build/worker.js
```

# C3 Generated Configuration

## Output Structure

```
my-app/
├── src/index.ts          # Worker entry point
├── wrangler.jsonc        # Cloudflare config
├── package.json          # Scripts
├── tsconfig.json
└── .gitignore
```

## wrangler.jsonc

```jsonc
{
  "$schema": "https://raw.githubusercontent.com/cloudflare/workers-sdk/main/packages/wrangler/config-schema.json",
  "name": "my-app",
  "main": "src/index.ts",
  "compatibility_date": "2026-01-27"
}
```

## Binding Placeholders

C3 generates **placeholder IDs** that must be replaced before deploy:

```jsonc
{
  "kv_namespaces": [{ "binding": "MY_KV", "id": "placeholder_kv_id" }],
  "d1_databases": [{ "binding": "DB", "database_id": "00000000-..." }]
}
```

**Replace with real IDs:**
```bash
npx wrangler kv namespace create MY_KV   # Returns real ID
npx wrangler d1 create my-database       # Returns real database_id
```

**Deployment error if not replaced:**
```
Error: Invalid KV namespace ID "placeholder_kv_id"
```

## Scripts

```json
{
  "scripts": {
    "dev": "wrangler dev",
    "deploy": "wrangler deploy",
    "cf-typegen": "wrangler types"
  }
}
```

## Type Generation

Run after adding bindings:
```bash
npm run cf-typegen
```

Generates `.wrangler/types/runtime.d.ts`:
```typescript
interface Env {
  MY_KV: KVNamespace;
  DB: D1Database;
}
```

## Post-Creation Checklist

1. Review `wrangler.jsonc` - check name, compatibility_date
2. Replace placeholder binding IDs with real resource IDs
3. Run `npm run cf-typegen`
4. Test: `npm run dev`
5. Deploy: `npm run deploy`
6. Add secrets: `npx wrangler secret put SECRET_NAME`

# C3 Troubleshooting

## Deployment Issues

### Placeholder IDs

**Error:** "Invalid namespace ID"  
**Fix:** Replace placeholders in wrangler.jsonc with real IDs:
```bash
npx wrangler kv namespace create MY_KV  # Get real ID
```

### Authentication

**Error:** "Not authenticated"  
**Fix:** `npx wrangler login` or set `CLOUDFLARE_API_TOKEN`

### Name Conflict

**Error:** "Worker already exists"  
**Fix:** Change `name` in wrangler.jsonc

## Platform Selection

| Need | Platform |
|------|----------|
| Git integration, branch previews | `--platform=pages` |
| Durable Objects, D1, Queues | Workers (default) |

Wrong platform? Recreate with correct `--platform` flag.

## TypeScript Issues

**"Cannot find name 'KVNamespace'"**
```bash
npm run cf-typegen  # Regenerate types
# Restart TS server in editor
```

**Missing types after config change:** Re-run `npm run cf-typegen`

## Package Manager

**Multiple lockfiles causing issues:**
```bash
rm pnpm-lock.yaml  # If using npm
rm package-lock.json  # If using pnpm
```

## CI/CD

**CI hangs on prompts:**
```bash
npm create cloudflare@latest my-app -- \
  --type=hello-world --lang=ts --no-git --no-deploy
```

**Auth in CI:**
```yaml
env:
  CLOUDFLARE_API_TOKEN: ${{ secrets.CLOUDFLARE_API_TOKEN }}
  CLOUDFLARE_ACCOUNT_ID: ${{ secrets.CLOUDFLARE_ACCOUNT_ID }}
```

## Framework-Specific

| Framework | Issue | Fix |
|-----------|-------|-----|
| Next.js | create-next-app failed | `npm cache clean --force`, retry |
| Astro | Adapter missing | Install `@astrojs/cloudflare` |
| Remix | Module errors | Update `@remix-run/cloudflare*` |

## Compatibility Date

**"Feature X requires compatibility_date >= ..."**  
**Fix:** Update `compatibility_date` in wrangler.jsonc to today's date

## Node.js Version

**"Node.js version not supported"**  
**Fix:** Install Node.js 18+ (`nvm install 20`)

## Quick Reference

| Error | Cause | Fix |
|-------|-------|-----|
| Invalid namespace ID | Placeholder binding | Create resource, update config |
| Not authenticated | No login | `npx wrangler login` |
| Cannot find KVNamespace | Missing types | `npm run cf-typegen` |
| Worker already exists | Name conflict | Change `name` |
| CI hangs | Missing flags | Add --type, --lang, --no-deploy |
| Template not found | Bad name | Check cloudflare/templates |

# C3 Usage Patterns

## Quick Workflows

For framework apps, follow [Framework Setup](README.md#framework-setup).

```bash
# TypeScript API Worker
npm create cloudflare@latest my-api -- --type=hello-world --lang=ts --deploy

# Astro static site  
npm create cloudflare@latest my-blog -- --type=web-app --framework=astro --platform=pages --ts
```

## CI/CD (GitHub Actions)

```yaml
- name: Deploy
  run: npm run deploy
  env:
    CLOUDFLARE_API_TOKEN: ${{ secrets.CLOUDFLARE_API_TOKEN }}
    CLOUDFLARE_ACCOUNT_ID: ${{ secrets.CLOUDFLARE_ACCOUNT_ID }}
```

**Non-interactive requires:**
```bash
--type=<value>       # Required
--no-git             # Recommended (CI already in git)
--no-deploy          # Deploy separately with secrets
--framework=<value>  # For web-app
--ts / --no-ts       # Required
```

## Monorepo

C3 detects workspace config (`package.json` workspaces or `pnpm-workspace.yaml`).

```bash
cd packages/
npm create cloudflare@latest my-worker -- --type=hello-world --lang=ts --no-deploy
```

## Custom Templates

```bash
# GitHub repo
npm create cloudflare@latest -- --template=username/repo
npm create cloudflare@latest -- --template=cloudflare/templates/worker-openapi

# Local path
npm create cloudflare@latest my-app -- --template=../my-template
```

**Template requires `c3.config.json`:**
```json
{
  "name": "my-template",
  "category": "hello-world",
  "copies": [{ "path": "src/" }, { "path": "wrangler.jsonc" }],
  "transforms": [{ "path": "package.json", "jsonc": { "name": "{{projectName}}" }}]
}
```

## Existing Projects

```bash
# Add Cloudflare to existing Worker
npm create cloudflare@latest . -- --type=pre-existing --existing-script=./dist/index.js
```

For existing framework apps, follow [Framework Setup](README.md#framework-setup).

## Post-Creation Checklist

1. Review `wrangler.jsonc` - set `compatibility_date`, verify `name`
2. Create bindings: `wrangler kv namespace create`, `wrangler d1 create`, `wrangler r2 bucket create`
3. Generate types: `npm run cf-typegen`
4. Test: `npm run dev`
5. Deploy: `npm run deploy`
6. Set secrets: `wrangler secret put SECRET_NAME`

# C3 (create-cloudflare)

Official CLI for scaffolding Cloudflare Workers and Pages projects with templates, TypeScript, and instant deployment.

## Quick Start

```bash
# Interactive (recommended for first-time)
npm create cloudflare@latest my-app

# Worker (API/WebSocket/Cron)
npm create cloudflare@latest my-api -- --type=hello-world --ts

# Pages (static/SSG)
npm create cloudflare@latest my-site -- --type=web-app --framework=astro --platform=pages
```

## Platform Decision Tree

```
What are you building?

├─ API / WebSocket / Cron / Email handler
│   └─ Workers (default) - no --platform flag needed
│       npm create cloudflare@latest my-api -- --type=hello-world

├─ Static site / SSG / Documentation
│   └─ Pages - requires --platform=pages
│       npm create cloudflare@latest my-site -- --type=web-app --framework=astro --platform=pages

├─ Full-stack app (Next.js/Remix/SvelteKit)
│   └─ Follow the current framework guide below

└─ Convert existing project
    └─ npm create cloudflare@latest . -- --type=pre-existing --existing-script=./src/worker.ts
```

**Critical:** Pages projects require `--platform=pages` flag. Without it, C3 defaults to Workers.

## Framework Setup

Fetch the [Workers framework guide](https://developers.cloudflare.com/workers/framework-guides/) for the chosen framework before scaffolding or adapting an existing app. For Next.js, follow [Next.js on Workers](https://developers.cloudflare.com/workers/framework-guides/web-apps/nextjs/); use the [Pages static export guide](https://developers.cloudflare.com/pages/framework-guides/nextjs/deploy-a-static-nextjs-site/) only when targeting a Next.js static export on Pages.

## Interactive Flow

When run without flags, C3 prompts in this order:

1. **Project name** - Directory to create (defaults to current dir with `.`)
2. **Application type** - `hello-world`, `web-app`, `demo`, `pre-existing`, `remote-template`
3. **Platform** - `workers` (default) or `pages` (for web apps only)
4. **Framework** - If web-app: `next`, `remix`, `astro`, `react-router`, `solid`, `svelte`, etc.
5. **TypeScript** - `yes` (recommended) or `no`
6. **Git** - Initialize repository? `yes` or `no`
7. **Deploy** - Deploy now? `yes` or `no` (requires `wrangler login`)

## Installation Methods

```bash
# NPM
npm create cloudflare@latest

# Yarn
yarn create cloudflare

# PNPM
pnpm create cloudflare@latest
```

## In This Reference

| File | Purpose | Use When |
|------|---------|----------|
| **api.md** | Complete CLI flag reference | Scripting, CI/CD, advanced usage |
| **configuration.md** | Generated files, bindings, types | Understanding output, customization |
| **patterns.md** | Workflows, CI/CD, monorepos | Real-world integration |
| **gotchas.md** | Troubleshooting failures | Deployment blocked, errors |

## Reading Order

| Task | Read |
|------|------|
| Create first project | README only |
| Set up CI/CD | README → api → patterns |
| Debug failed deploy | gotchas |
| Understand generated files | configuration |
| Full CLI reference | api |
| Create custom template | patterns → configuration |
| Convert existing project | README → patterns |

## Post-Creation

```bash
cd my-app

# Local dev with hot reload
npm run dev

# Generate TypeScript types for bindings
npm run cf-typegen

# Deploy to Cloudflare
npm run deploy
```

## See Also

- **workers/README.md** - Workers runtime, bindings, APIs
- **workers-ai/README.md** - AI/ML models
- **pages/README.md** - Pages-specific features
- **wrangler/README.md** - Wrangler CLI beyond initial setup
- **d1/README.md** - SQLite database
- **r2/README.md** - Object storage

# Cache Reserve API

## Workers Integration

```
┌────────────────────────────────────────────────────────────────┐
│ CRITICAL: Workers Cache API ≠ Cache Reserve                   │
│                                                                │
│ • Workers caches.default / cache.put() → edge cache ONLY      │
│ • Cache Reserve → zone-level setting, automatic, no per-req   │
│ • You CANNOT selectively write to Cache Reserve from Workers  │
│ • Cache Reserve works with standard fetch(), not cache.put()  │
└────────────────────────────────────────────────────────────────┘
```

Cache Reserve is a **zone-level configuration**, not a per-request API. It works automatically when enabled for the zone:

### Standard Fetch (Recommended)

```typescript
// Cache Reserve works automatically via standard fetch
export default {
  async fetch(request: Request, env: Env): Promise<Response> {
    // Standard fetch uses Cache Reserve automatically
    return await fetch(request);
  }
};
```

### Cache API Limitations

**IMPORTANT**: `cache.put()` is **NOT compatible** with Cache Reserve or Tiered Cache.

```typescript
// ❌ WRONG: cache.put() bypasses Cache Reserve
const cache = caches.default;
let response = await cache.match(request);
if (!response) {
  response = await fetch(request);
  await cache.put(request, response.clone()); // Bypasses Cache Reserve!
}

// ✅ CORRECT: Use standard fetch for Cache Reserve compatibility
return await fetch(request);

// ✅ CORRECT: Use Cache API only for custom cache namespaces
const customCache = await caches.open('my-custom-cache');
let response = await customCache.match(request);
if (!response) {
  response = await fetch(request);
  await customCache.put(request, response.clone()); // Custom cache OK
}
```

## Purging and Cache Management

### Purge by URL (Instant)

```typescript
// Purge specific URL from Cache Reserve immediately
const purgeCacheReserveByURL = async (
  zoneId: string,
  apiToken: string,
  urls: string[]
) => {
  const response = await fetch(
    `https://api.cloudflare.com/client/v4/zones/${zoneId}/purge_cache`,
    {
      method: 'POST',
      headers: {
        'Authorization': `Bearer ${apiToken}`,
        'Content-Type': 'application/json',
      },
      body: JSON.stringify({ files: urls })
    }
  );
  return await response.json();
};

// Example usage
await purgeCacheReserveByURL('zone123', 'token456', [
  'https://example.com/image.jpg',
  'https://example.com/video.mp4'
]);
```

### Purge by Tag/Host/Prefix (Revalidation)

```typescript
// Purge by cache tag - forces revalidation, not immediate removal
await fetch(
  `https://api.cloudflare.com/client/v4/zones/${zoneId}/purge_cache`,
  {
    method: 'POST',
    headers: { 'Authorization': `Bearer ${apiToken}`, 'Content-Type': 'application/json' },
    body: JSON.stringify({ tags: ['tag1', 'tag2'] })
  }
);
```

**Purge behavior:**
- **By URL**: Immediate removal from Cache Reserve + edge cache
- **By tag/host/prefix**: Revalidation only, assets remain in storage (costs continue)

### Clear All Cache Reserve Data

```typescript
// Requires Cache Reserve OFF first
await fetch(
  `https://api.cloudflare.com/client/v4/zones/${zoneId}/cache/cache_reserve_clear`,
  { method: 'POST', headers: { 'Authorization': `Bearer ${apiToken}` } }
);

// Check status: GET same endpoint returns { state: "In-progress" | "Completed" }
```

**Process**: Disable Cache Reserve → Call clear endpoint → Wait up to 24hr → Re-enable

## Monitoring and Analytics

### Dashboard Analytics

Navigate to **Caching > Cache Reserve** to view:

- **Egress Savings**: Total bytes served from Cache Reserve vs origin egress cost saved
- **Requests Served**: Cache Reserve hits vs misses breakdown
- **Storage Used**: Current GB stored in Cache Reserve (billed monthly)
- **Operations**: Class A (writes) and Class B (reads) operation counts
- **Cost Tracking**: Estimated monthly costs based on current usage

### Logpush Integration

```typescript
// Logpush field: CacheReserveUsed (boolean) - filter for Cache Reserve hits
// Query Cache Reserve hits in analytics
const logpushQuery = `
  SELECT 
    ClientRequestHost, 
    COUNT(*) as requests, 
    SUM(EdgeResponseBytes) as bytes_served,
    COUNT(CASE WHEN CacheReserveUsed = true THEN 1 END) as cache_reserve_hits,
    COUNT(CASE WHEN CacheReserveUsed = false THEN 1 END) as cache_reserve_misses
  FROM http_requests 
  WHERE Timestamp >= NOW() - INTERVAL '24 hours'
  GROUP BY ClientRequestHost 
  ORDER BY requests DESC
`;

// Filter only Cache Reserve hits
const crHitsQuery = `
  SELECT ClientRequestHost, COUNT(*) as requests, SUM(EdgeResponseBytes) as bytes
  FROM http_requests 
  WHERE CacheReserveUsed = true AND Timestamp >= NOW() - INTERVAL '7 days'
  GROUP BY ClientRequestHost 
  ORDER BY bytes DESC
`;
```

### GraphQL Analytics

```graphql
query CacheReserveAnalytics($zoneTag: string, $since: string, $until: string) {
  viewer {
    zones(filter: { zoneTag: $zoneTag }) {
      httpRequests1dGroups(
        filter: { datetime_geq: $since, datetime_leq: $until }
        limit: 1000
      ) {
        dimensions { date }
        sum {
          cachedBytes
          cachedRequests
          bytes
          requests
        }
      }
    }
  }
}
```

## Pricing

```typescript
// Storage: $0.015/GB-month | Class A (writes): $4.50/M | Class B (reads): $0.36/M
// Cache miss: 1A + 1B | Cache hit: 1B | Assets >1GB: proportionally more ops
```

## See Also

- [README](./README.md) - Overview and core concepts
- [Configuration](./configuration.md) - Setup and Cache Rules
- [Patterns](./patterns.md) - Best practices and optimization
- [Gotchas](./gotchas.md) - Common issues and troubleshooting

# Cache Reserve Configuration

## Dashboard Setup

**Minimum steps to enable:**

```bash
# Navigate to dashboard
https://dash.cloudflare.com/caching/cache-reserve

# Click "Enable Storage Sync" or "Purchase" button
```

**Prerequisites:**
- Paid Cache Reserve plan or Smart Shield Advanced required
- Tiered Cache **required** for Cache Reserve to function optimally

## API Configuration

### REST API

```bash
# Enable
curl -X PATCH "https://api.cloudflare.com/client/v4/zones/$ZONE_ID/cache/cache_reserve" \
  -H "Authorization: Bearer $API_TOKEN" -H "Content-Type: application/json" \
  -d '{"value": "on"}'

# Check status
curl -X GET "https://api.cloudflare.com/client/v4/zones/$ZONE_ID/cache/cache_reserve" \
  -H "Authorization: Bearer $API_TOKEN"
```

### TypeScript SDK

```bash
npm install cloudflare
```

```typescript
import Cloudflare from 'cloudflare';

const client = new Cloudflare({
  apiToken: process.env.CLOUDFLARE_API_TOKEN,
});

// Enable Cache Reserve
await client.cache.cacheReserve.edit({
  zone_id: 'abc123',
  value: 'on',
});

// Get Cache Reserve status
const status = await client.cache.cacheReserve.get({
  zone_id: 'abc123',
});
console.log(status.value); // 'on' or 'off'
```

### Python SDK

```bash
pip install cloudflare
```

```python
from cloudflare import Cloudflare

client = Cloudflare(api_token=os.environ.get("CLOUDFLARE_API_TOKEN"))

# Enable Cache Reserve
client.cache.cache_reserve.edit(
    zone_id="abc123",
    value="on"
)

# Get Cache Reserve status
status = client.cache.cache_reserve.get(zone_id="abc123")
print(status.value)  # 'on' or 'off'
```

### Terraform

```hcl
terraform {
  required_providers {
    cloudflare = {
      source  = "cloudflare/cloudflare"
      version = "~> 4.0"
    }
  }
}

provider "cloudflare" {
  api_token = var.cloudflare_api_token
}

resource "cloudflare_zone_cache_reserve" "example" {
  zone_id = var.zone_id
  enabled = true
}

# Tiered Cache is required for Cache Reserve
resource "cloudflare_tiered_cache" "example" {
  zone_id    = var.zone_id
  cache_type = "smart"
}
```

### Pulumi

```typescript
import * as cloudflare from "@pulumi/cloudflare";

// Enable Cache Reserve
const cacheReserve = new cloudflare.ZoneCacheReserve("example", {
  zoneId: zoneId,
  enabled: true,
});

// Enable Tiered Cache (required)
const tieredCache = new cloudflare.TieredCache("example", {
  zoneId: zoneId,
  cacheType: "smart",
});
```

### Required API Token Permissions

- `Zone Settings Read`
- `Zone Settings Write`
- `Zone Read`
- `Zone Write`

## Cache Rules Integration

Control Cache Reserve eligibility via Cache Rules:

```typescript
// Enable for static assets
{
  action: 'set_cache_settings',
  action_parameters: {
    cache_reserve: { eligible: true, minimum_file_ttl: 86400 },
    edge_ttl: { mode: 'override_origin', default: 86400 },
    cache: true
  },
  expression: '(http.request.uri.path matches "\\.(jpg|png|webp|pdf|zip)$")'
}

// Disable for APIs
{
  action: 'set_cache_settings',
  action_parameters: { cache_reserve: { eligible: false } },
  expression: '(http.request.uri.path matches "^/api/")'
}

// Create via API: PUT to zones/{zone_id}/rulesets/phases/http_request_cache_settings/entrypoint
```

## Wrangler Integration

Cache Reserve works automatically with Workers deployed via Wrangler. No special wrangler.jsonc configuration needed - enable Cache Reserve via Dashboard or API for the zone.

## See Also

- [README](./README.md) - Overview and core concepts
- [API Reference](./api.md) - Purging and monitoring APIs
- [Patterns](./patterns.md) - Best practices and optimization
- [Gotchas](./gotchas.md) - Common issues and troubleshooting

# Cache Reserve Gotchas

## Common Errors

### "Assets Not Being Cached in Cache Reserve"

**Cause:** Asset is not cacheable, TTL < 10 hours, Content-Length header missing, or blocking headers present (Set-Cookie, Vary: *)  
**Solution:** Ensure minimum TTL of 10+ hours (`Cache-Control: public, max-age=36000`), add Content-Length header, remove Set-Cookie header, and set `Vary: Accept-Encoding` (not *)

### "Range Requests Not Working" (Video Seeking Fails)

**Cause:** Cache Reserve does **NOT** support range requests (HTTP 206 Partial Content)  
**Solution:** Range requests bypass Cache Reserve entirely. For video streaming with seeking:
- Use edge cache only (shorter TTLs)
- Consider R2 with direct access for range-heavy workloads
- Accept that seekable content won't benefit from Cache Reserve persistence

### "Origin Bandwidth Higher Than Expected"

**Cause:** Cache Reserve fetches **uncompressed** content from origin, even though it serves compressed to visitors  
**Solution:** 
- If origin charges by bandwidth, factor in uncompressed transfer costs
- Cache Reserve compresses for visitors automatically (saves visitor bandwidth)
- Compare: origin egress savings vs higher uncompressed fetch costs

### "Cloudflare Images Not Caching with Cache Reserve"

**Cause:** Cloudflare Images with `Vary: Accept` header (format negotiation) is incompatible with Cache Reserve  
**Solution:** 
- Cache Reserve silently skips images with Vary for format negotiation
- Original images (non-transformed) may still be eligible
- Use Cloudflare Images variants or edge cache for transformed images

### "High Class A Operations Costs"

**Cause:** Frequent cache misses, short TTLs, or frequent revalidation  
**Solution:** Increase TTL for stable content (24+ hours), enable Tiered Cache to reduce direct Cache Reserve misses, or use stale-while-revalidate

### "Purge Not Working as Expected"

**Cause:** Purge by tag only triggers revalidation but doesn't remove from Cache Reserve storage  
**Solution:** Use purge by URL for immediate removal, or disable Cache Reserve then clear all data for complete removal

### "O2O (Orange-to-Orange) Assets Not Caching"

**Cause:** Orange-to-Orange (proxied zone requesting another proxied zone on Cloudflare) bypasses Cache Reserve  
**Solution:** 
- **What is O2O**: Zone A (proxied) → Zone B (proxied), both on Cloudflare
- **Detection**: Check `cf-cache-status` for `BYPASS` and review request path
- **Workaround**: Use R2 or direct origin access instead of O2O proxy chains

### "Cache Reserve must be OFF before clearing data"

**Cause:** Attempting to clear Cache Reserve data while it's still enabled  
**Solution:** Disable Cache Reserve first, wait briefly for propagation (5s), then clear data (can take up to 24 hours)

## Limits

| Limit | Value | Notes |
|-------|-------|-------|
| Minimum TTL | 10 hours (36000 seconds) | Assets with shorter TTL not eligible |
| Default retention | 30 days (2592000 seconds) | Configurable |
| Maximum file size | Same as R2 limits | No practical limit |
| Purge/clear time | Up to 24 hours | Complete propagation time |
| Plan requirement | Paid Cache Reserve or Smart Shield | Not available on free plans |
| Content-Length header | Required | Must be present for eligibility |
| Set-Cookie header | Blocks caching | Must not be present (or use private directive) |
| Vary header | Cannot be * | Can use Vary: Accept-Encoding |
| Image transformations | Variants not eligible | Original images only |
| Range requests | NOT supported | HTTP 206 bypasses Cache Reserve |
| Compression | Fetches uncompressed | Serves compressed to visitors |
| Worker control | Zone-level only | Cannot control per-request |
| O2O requests | Bypassed | Orange-to-Orange not eligible |

## Additional Resources

- **Official Docs**: https://developers.cloudflare.com/cache/advanced-configuration/cache-reserve/
- **API Reference**: https://developers.cloudflare.com/api/resources/cache/subresources/cache_reserve/
- **Cache Rules**: https://developers.cloudflare.com/cache/how-to/cache-rules/
- **Workers Cache API**: https://developers.cloudflare.com/workers/runtime-apis/cache/
- **R2 Documentation**: https://developers.cloudflare.com/r2/
- **Smart Shield**: https://developers.cloudflare.com/smart-shield/
- **Tiered Cache**: https://developers.cloudflare.com/cache/how-to/tiered-cache/

## Troubleshooting Flowchart

Asset not caching in Cache Reserve?

```
1. Is Cache Reserve enabled for zone?
   → No: Enable via Dashboard or API
   → Yes: Continue to step 2

2. Is Tiered Cache enabled?
   → No: Enable Tiered Cache (required!)
   → Yes: Continue to step 3

3. Does asset have TTL ≥ 10 hours?
   → No: Increase via Cache Rules (edge_ttl override)
   → Yes: Continue to step 4

4. Is Content-Length header present?
   → No: Fix origin to include Content-Length
   → Yes: Continue to step 5

5. Is Set-Cookie header present?
   → Yes: Remove Set-Cookie or scope appropriately
   → No: Continue to step 6

6. Is Vary header set to *?
   → Yes: Change to specific value (e.g., Accept-Encoding)
   → No: Continue to step 7

7. Is this a range request?
   → Yes: Range requests bypass Cache Reserve (not supported)
   → No: Continue to step 8

8. Is this an O2O (Orange-to-Orange) request?
   → Yes: O2O bypasses Cache Reserve
   → No: Continue to step 9

9. Check Logpush CacheReserveUsed field
   → Filter logs to see if assets ever hit Cache Reserve
   → Verify cf-cache-status header (should be HIT after first request)
```

## See Also

- [README](./README.md) - Overview and core concepts
- [Configuration](./configuration.md) - Setup and Cache Rules
- [API Reference](./api.md) - Purging and monitoring
- [Patterns](./patterns.md) - Best practices and optimization

# Cache Reserve Patterns

## Best Practices

### 1. Always Enable Tiered Cache

```typescript
// Cache Reserve is designed for use WITH Tiered Cache
const configuration = {
  tieredCache: 'enabled',    // Required for optimal performance
  cacheReserve: 'enabled',   // Works best with Tiered Cache
  
  hierarchy: [
    'Lower-Tier Cache (visitor)',
    'Upper-Tier Cache (origin region)',
    'Cache Reserve (persistent)',
    'Origin'
  ]
};
```

### 2. Set Appropriate Cache-Control Headers

```typescript
// Origin response headers for Cache Reserve eligibility
const originHeaders = {
  'Cache-Control': 'public, max-age=86400', // 24hr (minimum 10hr)
  'Content-Length': '1024000', // Required
  'Cache-Tag': 'images,product-123', // Optional: purging
  'ETag': '"abc123"', // Optional: revalidation
  // Avoid: 'Set-Cookie' and 'Vary: *' prevent caching
};
```

### 3. Use Cache Rules for Fine-Grained Control

```typescript
// Different TTLs for different content types
const cacheRules = [
  {
    description: 'Long-term cache for immutable assets',
    expression: '(http.request.uri.path matches "^/static/.*\\.[a-f0-9]{8}\\.")',
    action_parameters: {
      cache_reserve: { eligible: true },
      edge_ttl: { mode: 'override_origin', default: 2592000 }, // 30 days
      cache: true
    }
  },
  {
    description: 'Moderate cache for regular images',
    expression: '(http.request.uri.path matches "\\.(jpg|png|webp)$")',
    action_parameters: {
      cache_reserve: { eligible: true },
      edge_ttl: { mode: 'override_origin', default: 86400 }, // 24 hours
      cache: true
    }
  },
  {
    description: 'Exclude API from Cache Reserve',
    expression: '(http.request.uri.path matches "^/api/")',
    action_parameters: { cache_reserve: { eligible: false }, cache: false }
  }
];
```

### 4. Making Assets Cache Reserve Eligible from Workers

**Note**: This modifies response headers to meet eligibility criteria but does NOT directly control Cache Reserve storage (which is zone-level automatic).

```typescript
export default {
  async fetch(request: Request, env: Env): Promise<Response> {
    const response = await fetch(request);
    if (!response.ok) return response;
    
    const headers = new Headers(response.headers);
    headers.set('Cache-Control', 'public, max-age=36000'); // 10hr minimum
    headers.delete('Set-Cookie'); // Blocks caching
    
    // Ensure Content-Length present
    if (!headers.has('Content-Length')) {
      const blob = await response.blob();
      headers.set('Content-Length', blob.size.toString());
      return new Response(blob, { status: response.status, headers });
    }
    
    return new Response(response.body, { status: response.status, headers });
  }
};
```

### 5. Hostname Best Practices

Use Worker's hostname for efficient caching - avoid overriding hostname unnecessarily.

## Architecture Patterns

### Multi-Tier Caching + Immutable Assets

```typescript
// Optimal: L1 (visitor) → L2 (region) → L3 (Cache Reserve) → Origin
export default {
  async fetch(request: Request, env: Env): Promise<Response> {
    const url = new URL(request.url);
    const isImmutable = /\.[a-f0-9]{8,}\.(js|css|jpg|png|woff2)$/.test(url.pathname);
    const response = await fetch(request);
    
    if (isImmutable) {
      const headers = new Headers(response.headers);
      headers.set('Cache-Control', 'public, max-age=31536000, immutable');
      return new Response(response.body, { status: response.status, headers });
    }
    return response;
  }
};
```

## Cost Optimization

### Cost Calculator

```typescript
interface CacheReserveEstimate {
  avgAssetSizeGB: number;
  uniqueAssets: number;
  monthlyReads: number;
  monthlyWrites: number;
  originEgressCostPerGB: number; // e.g., AWS: $0.09/GB
}

function estimateMonthlyCost(input: CacheReserveEstimate) {
  // Cache Reserve pricing
  const storageCostPerGBMonth = 0.015;
  const classAPerMillion = 4.50; // writes
  const classBPerMillion = 0.36; // reads
  
  // Calculate Cache Reserve costs
  const totalStorageGB = input.avgAssetSizeGB * input.uniqueAssets;
  const storageCost = totalStorageGB * storageCostPerGBMonth;
  const writeCost = (input.monthlyWrites / 1_000_000) * classAPerMillion;
  const readCost = (input.monthlyReads / 1_000_000) * classBPerMillion;
  
  const cacheReserveCost = storageCost + writeCost + readCost;
  
  // Calculate origin egress cost (what you'd pay without Cache Reserve)
  const totalTrafficGB = (input.monthlyReads * input.avgAssetSizeGB);
  const originEgressCost = totalTrafficGB * input.originEgressCostPerGB;
  
  // Savings calculation
  const savings = originEgressCost - cacheReserveCost;
  const savingsPercent = ((savings / originEgressCost) * 100).toFixed(1);
  
  return {
    cacheReserveCost: `$${cacheReserveCost.toFixed(2)}`,
    originEgressCost: `$${originEgressCost.toFixed(2)}`,
    monthlySavings: `$${savings.toFixed(2)}`,
    savingsPercent: `${savingsPercent}%`,
    breakdown: {
      storage: `$${storageCost.toFixed(2)}`,
      writes: `$${writeCost.toFixed(2)}`,
      reads: `$${readCost.toFixed(2)}`,
    }
  };
}

// Example: Media library
const mediaLibrary = estimateMonthlyCost({
  avgAssetSizeGB: 0.005, // 5MB images
  uniqueAssets: 10_000,
  monthlyReads: 5_000_000,
  monthlyWrites: 50_000,
  originEgressCostPerGB: 0.09, // AWS S3
});

console.log(mediaLibrary);
// {
//   cacheReserveCost: "$9.98",
//   originEgressCost: "$25.00",
//   monthlySavings: "$15.02",
//   savingsPercent: "60.1%",
//   breakdown: { storage: "$0.75", writes: "$0.23", reads: "$9.00" }
// }
```

### Optimization Guidelines

- **Set appropriate TTLs**: 10hr minimum, 24hr+ optimal for stable content, 30d max cautiously
- **Cache high-value stable assets**: Images, media, fonts, archives, documentation
- **Exclude frequently changing**: APIs, user-specific content, real-time data
- **Compression note**: Cache Reserve fetches uncompressed from origin, serves compressed to visitors - factor in origin egress costs

## See Also

- [README](./README.md) - Overview and core concepts
- [Configuration](./configuration.md) - Setup and Cache Rules
- [API Reference](./api.md) - Purging and monitoring
- [Gotchas](./gotchas.md) - Common issues and troubleshooting

# Cloudflare Cache Reserve

**Persistent cache storage built on R2 for long-term content retention**

## Smart Shield Integration

Cache Reserve is part of **Smart Shield**, Cloudflare's comprehensive security and performance suite:

- **Smart Shield Advanced tier**: Includes 2TB Cache Reserve storage
- **Standalone purchase**: Available separately if not using Smart Shield
- **Migration**: Existing standalone customers can migrate to Smart Shield bundles

**Decision**: Already on Smart Shield Advanced? Cache Reserve is included. Otherwise evaluate standalone purchase vs Smart Shield upgrade.

## Overview

Cache Reserve is Cloudflare's persistent, large-scale cache storage layer built on R2. It acts as the ultimate upper-tier cache, storing cacheable content for extended periods (30+ days) to maximize cache hits, reduce origin egress fees, and shield origins from repeated requests for long-tail content.

## Core Concepts

### What is Cache Reserve?

- **Persistent storage layer**: Built on R2, sits above tiered cache hierarchy
- **Long-term retention**: 30-day default retention, extended on each access
- **Automatic operation**: Works seamlessly with existing CDN, no code changes required
- **Origin shielding**: Dramatically reduces origin egress by serving cached content longer
- **Usage-based pricing**: Pay only for storage + read/write operations

### Cache Hierarchy

```
Visitor Request
    ↓
Lower-Tier Cache (closest to visitor)
    ↓ (on miss)
Upper-Tier Cache (closest to origin)
    ↓ (on miss)
Cache Reserve (R2 persistent storage)
    ↓ (on miss)
Origin Server
```

### How It Works

1. **On cache miss**: Content fetched from origin �� written to Cache Reserve + edge caches simultaneously
2. **On edge eviction**: Content may be evicted from edge cache but remains in Cache Reserve
3. **On subsequent request**: If edge cache misses but Cache Reserve hits → content restored to edge caches
4. **Retention**: Assets remain in Cache Reserve for 30 days since last access (configurable via TTL)

## When to Use Cache Reserve

```
Need persistent caching?
├─ High origin egress costs → Cache Reserve ✓
├─ Long-tail content (archives, media libraries) → Cache Reserve ✓
├─ Already using Smart Shield Advanced → Included! ✓
├─ Video streaming with seeking (range requests) → ✗ Not supported
├─ Dynamic/personalized content → ✗ Use edge cache only
├─ Need per-request cache control from Workers → ✗ Use R2 directly
└─ Frequently updated content (< 10hr lifetime) → ✗ Not eligible
```

## Asset Eligibility

Cache Reserve only stores assets meeting **ALL** criteria:

- Cacheable per Cloudflare's standard rules
- Minimum 10-hour TTL (36000 seconds)
- `Content-Length` header present
- Original files only (not transformed images)

### Eligibility Checklist

Use this checklist to verify if an asset is eligible:

- [ ] Zone has Cache Reserve enabled
- [ ] Zone has Tiered Cache enabled (required)
- [ ] Asset TTL ≥ 10 hours (36,000 seconds)
- [ ] `Content-Length` header present on origin response
- [ ] No `Set-Cookie` header (or uses private directive)
- [ ] `Vary` header is NOT `*` (can be `Accept-Encoding`)
- [ ] Not an image transformation variant (original images OK)
- [ ] Not a range request (no HTTP 206 support)
- [ ] Not O2O (Orange-to-Orange) proxied request

**All boxes must be checked for Cache Reserve eligibility.**

### Not Eligible

- Assets with TTL < 10 hours
- Responses without `Content-Length` header
- Image transformation variants (original images are eligible)
- Responses with `Set-Cookie` headers
- Responses with `Vary: *` header
- Assets from R2 public buckets on same zone
- O2O (Orange-to-Orange) setup requests
- **Range requests** (video seeking, partial content downloads)

## Quick Start

```bash
# Enable via Dashboard
https://dash.cloudflare.com/caching/cache-reserve
# Click "Enable Storage Sync" or "Purchase" button
```

**Prerequisites:**
- Paid Cache Reserve plan or Smart Shield Advanced required
- Tiered Cache required for optimal performance

## Essential Commands

```bash
# Check Cache Reserve status
curl -X GET "https://api.cloudflare.com/client/v4/zones/$ZONE_ID/cache/cache_reserve" \
  -H "Authorization: Bearer $API_TOKEN"

# Enable Cache Reserve
curl -X PATCH "https://api.cloudflare.com/client/v4/zones/$ZONE_ID/cache/cache_reserve" \
  -H "Authorization: Bearer $API_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"value": "on"}'

# Check asset cache status
curl -I https://example.com/asset.jpg | grep -i cache
```

## In This Reference

| Task | Files |
|------|-------|
| Evaluate if Cache Reserve fits your use case | README.md (this file) |
| Enable Cache Reserve for your zone | README.md + [configuration.md](./configuration.md) |
| Use with Workers (understand limitations) | [api.md](./api.md) |
| Setup via SDKs or IaC (TypeScript, Python, Terraform) | [configuration.md](./configuration.md) |
| Optimize costs and debug issues | [patterns.md](./patterns.md) + [gotchas.md](./gotchas.md) |
| Understand eligibility and troubleshoot | [gotchas.md](./gotchas.md) → [patterns.md](./patterns.md) |

**Files:**
- [configuration.md](./configuration.md) - Setup, API, SDKs, and Cache Rules
- [api.md](./api.md) - Purging, monitoring, Workers integration
- [patterns.md](./patterns.md) - Best practices, cost optimization, debugging
- [gotchas.md](./gotchas.md) - Common issues, limitations, troubleshooting

## See Also
- [r2](../r2/) - Cache Reserve built on R2 storage
- [workers](https://developers.cloudflare.com/workers/) - Workers integration with Cache API

# Containers API

Use the [Container interface](https://developers.cloudflare.com/containers/reference/container-class/) for the current SDK methods, signatures, properties, and examples. Use the [Durable Object Container API](https://developers.cloudflare.com/durable-objects/api/container/) when working directly with the runtime rather than the SDK class.

| Task | Documentation |
| --- | --- |
| Forward HTTP or WebSocket requests | [Request methods](https://developers.cloudflare.com/containers/reference/container-class/#request-methods) |
| Start a process, wait for ports, stop, or destroy it | [Start and stop](https://developers.cloudflare.com/containers/reference/container-class/#start-and-stop) |
| React to startup, exit, errors, or idle expiry | [Lifecycle hooks](https://developers.cloudflare.com/containers/reference/container-class/#lifecycle-hooks) |
| Inspect state or keep background work active | [State and monitoring](https://developers.cloudflare.com/containers/reference/container-class/#state-and-monitoring) |
| Schedule callbacks without replacing the SDK's alarm handler | [Scheduling](https://developers.cloudflare.com/containers/reference/container-class/#scheduling) |
| Address named instances, select stateless instances, or switch request ports | [Utility functions](https://developers.cloudflare.com/containers/reference/container-class/#utility-functions) |
| Communicate over TCP from a Durable Object | [TCP port API](https://developers.cloudflare.com/durable-objects/api/container/#gettcpport) |

# Containers configuration

Read the relevant documentation before choosing configuration fields or resource sizes.

| Task | Documentation |
| --- | --- |
| Configure the container image, Durable Object binding, class, migrations, and instance count | [Wrangler Containers configuration](https://developers.cloudflare.com/workers/wrangler/configuration/#containers) |
| Select a predefined size or configure custom CPU, memory, and disk | [Limits and instance types](https://developers.cloudflare.com/containers/platform/limits/) and [custom instance configuration](https://developers.cloudflare.com/workers/wrangler/configuration/#custom-instance-types) |
| Set ports, readiness checks, idle timeout, entrypoint, or internet access | [Container properties](https://developers.cloudflare.com/containers/reference/container-class/#properties) |
| Set runtime variables or pass secrets per instance | [Environment variables](https://developers.cloudflare.com/containers/configuration/environment-variables/) and [environment variables and secrets example](https://developers.cloudflare.com/containers/examples/env-vars-and-secrets/) |
| Build images or use existing registry images | [Image management](https://developers.cloudflare.com/containers/guides/image-management/) |
| Run and iterate locally | [Local development](https://developers.cloudflare.com/containers/guides/local-dev/) |
| Deploy from a workstation or Workers Builds | [Deploy Containers](https://developers.cloudflare.com/containers/guides/deploy/) |
| Control image updates and replacement of running instances | [Rollouts](https://developers.cloudflare.com/containers/configuration/rollouts/) |
| Estimate resource and network costs | [Pricing](https://developers.cloudflare.com/containers/platform/pricing/) |

# Containers troubleshooting

Use the current documentation to diagnose behavior instead of relying on copied timeout values, resource limits, or lifecycle recipes.

| Symptom or concern | Documentation to read |
| --- | --- |
| Startup timeout or unavailable port | [Start and stop](https://developers.cloudflare.com/containers/reference/container-class/#start-and-stop), [Container properties](https://developers.cloudflare.com/containers/reference/container-class/#properties), and [first-deploy provisioning](https://developers.cloudflare.com/containers/get-started/) |
| WebSocket forwarding fails | [WebSocket example](https://developers.cloudflare.com/containers/examples/websocket/) and [request methods](https://developers.cloudflare.com/containers/reference/container-class/#request-methods) |
| Background work stops on idle expiry | [Activity renewal](https://developers.cloudflare.com/containers/reference/container-class/#renewactivitytimeout) and [idle expiry hook](https://developers.cloudflare.com/containers/reference/container-class/#onactivityexpired) |
| Scheduled callbacks do not run | [Scheduling and alarm ownership](https://developers.cloudflare.com/containers/reference/container-class/#scheduling) |
| Shutdown cleanup or filesystem data loss | [Container shutdown and disk lifecycle](https://developers.cloudflare.com/containers/concepts/architecture/#container-shutdown) |
| Out-of-memory errors or resource exhaustion | [FAQ](https://developers.cloudflare.com/containers/faq/) and [limits and instance types](https://developers.cloudflare.com/containers/platform/limits/) |
| Instance count exceeded or unexpected request distribution | [Wrangler configuration](https://developers.cloudflare.com/workers/wrangler/configuration/#containers) and [scaling and routing](https://developers.cloudflare.com/containers/configuration/scaling-and-routing/) |
| Worker and container image versions differ after deployment | [Deployment behavior](https://developers.cloudflare.com/containers/guides/deploy/) and [rollouts](https://developers.cloudflare.com/containers/configuration/rollouts/) |
| Local behavior differs from deployed behavior | [Local development](https://developers.cloudflare.com/containers/guides/local-dev/) |
| Logs, cold starts, or runtime availability questions | [FAQ](https://developers.cloudflare.com/containers/faq/) |

# Containers patterns

Choose instance identity based on the workload: per-user/session or per-job identities for affinity, one shared identity for a singleton, and interchangeable instances for stateless requests. Read [scaling and routing](https://developers.cloudflare.com/containers/configuration/scaling-and-routing/) for current helpers and scaling behavior before implementing that choice.

| Task | Documentation |
| --- | --- |
| Distribute requests across stateless instances | [Stateless instances example](https://developers.cloudflare.com/containers/examples/stateless/) |
| Forward WebSocket connections | [WebSocket example](https://developers.cloudflare.com/containers/examples/websocket/) |
| React to lifecycle changes | [Status hooks example](https://developers.cloudflare.com/containers/examples/status-hooks/) |
| Handle shutdown and persist data across restarts | [Container lifecycle](https://developers.cloudflare.com/containers/concepts/architecture/) and [Container interface](https://developers.cloudflare.com/containers/reference/container-class/) |
| Keep long operations active or schedule callbacks | [Activity renewal](https://developers.cloudflare.com/containers/reference/container-class/#renewactivitytimeout) and [scheduling](https://developers.cloudflare.com/containers/reference/container-class/#scheduling) |
| Start containers on a cron schedule | [Cron container example](https://developers.cloudflare.com/containers/examples/cron/) |
| Route requests to multiple ports | [Request methods](https://developers.cloudflare.com/containers/reference/container-class/#request-methods) and [utility functions](https://developers.cloudflare.com/containers/reference/container-class/#utility-functions) |
| Access Workers bindings from the container | [Connect to Workers and bindings](https://developers.cloudflare.com/containers/configuration/workers-connections/) |

## Workflows and Queues

For multi-step orchestration, combine the [Workflows Workers API](https://developers.cloudflare.com/workflows/build/workers-api/) with the [Container API](api.md). For queue-driven jobs, read the [Queues consumer API](https://developers.cloudflare.com/queues/configuration/javascript-apis/#consumer) and [acknowledgement and retry behavior](https://developers.cloudflare.com/queues/configuration/batching-retries/#explicit-acknowledgement-and-retries) alongside the Container API. These pages document the component APIs; they are not end-to-end Container integration examples.

# Cloudflare Containers

Use this reference for containerized applications on the Workers platform, including container-enabled Durable Objects, lifecycle management, and request routing.

## Choose the runtime

Use [Containers](https://developers.cloudflare.com/containers/) for existing container images, custom runtimes, system dependencies, full filesystem access, or workloads needing additional CPU and memory. Use [Workers](https://developers.cloudflare.com/workers/) when the application fits the Workers runtime without those requirements.

Containers are controlled through [Durable Objects](https://developers.cloudflare.com/durable-objects/). An instance's identity does not make its filesystem persistent: design for restarts and store durable data outside the container disk. Read [Container lifecycle](https://developers.cloudflare.com/containers/concepts/architecture/) and [Container interface](https://developers.cloudflare.com/containers/reference/container-class/) for the relationship between the process, its Durable Object, and persistent storage.

## Find the documentation for the task

Read the linked page before writing code or configuration; use its current API, examples, and constraints rather than reconstructing them from memory.

| Task | Start here |
| --- | --- |
| Create a project and deploy the first container | [Get started](https://developers.cloudflare.com/containers/get-started/) |
| Configure images, bindings, instance sizes, and deployments | [Configuration](configuration.md) |
| Control startup, requests, lifecycle, and scheduling | [API](api.md) |
| Choose routing or connect other services | [Patterns](patterns.md) |
| Diagnose startup, persistence, capacity, or rollout issues | [Gotchas](gotchas.md) |

For additional topics, consult the [Containers documentation index](https://developers.cloudflare.com/containers/llms.txt).

# Cron Triggers API

Fetch the handler documentation before writing code; use its current language examples and completion semantics.

| Task | Documentation |
| --- | --- |
| Implement the handler and access the cron expression, scheduled time, bindings, and context | [Scheduled handler](https://developers.cloudflare.com/workers/runtime-apis/handlers/scheduled/) |
| Route different schedules to different operations | [Handle multiple cron triggers](https://developers.cloudflare.com/workers/runtime-apis/handlers/scheduled/#handle-multiple-cron-triggers) |
| Await work and understand how asynchronous failures affect invocation status | [Handler methods](https://developers.cloudflare.com/workers/runtime-apis/handlers/scheduled/#methods) |
| Invoke a scheduled handler locally with a chosen expression or time | [Test Cron Triggers locally](https://developers.cloudflare.com/workers/configuration/cron-triggers/#test-cron-triggers-locally) |
| Build tests using runtime-backed controllers and execution contexts | [Workers test APIs](https://developers.cloudflare.com/workers/testing/vitest-integration/test-apis/) |
| Start and inspect a Workflow instance | [Trigger Workflows](https://developers.cloudflare.com/workflows/build/trigger-workflows/) |

Decide which operation establishes successful completion and make its failures observable. Test each configured schedule and partial-failure recovery. Read the local-testing documentation for the current endpoint and query parameters instead of adding a production HTTP route to imitate the development helper.

See [patterns.md](./patterns.md) for execution design and [gotchas.md](./gotchas.md) for failures.

# Cron Triggers Configuration

Fetch the documentation for the configuration operation before changing schedules.

| Task | Documentation |
| --- | --- |
| Add a handler and configure triggers, including per-environment schedules and deployment propagation | [Add a Cron Trigger](https://developers.cloudflare.com/workers/configuration/cron-triggers/#add-a-cron-trigger) |
| Choose an expression, interpret weekday numbering, or check supported extensions | [Supported cron expressions](https://developers.cloudflare.com/workers/configuration/cron-triggers/#supported-cron-expressions) |
| Remove schedules or distinguish omission from an empty configuration | [Remove a Cron Trigger](https://developers.cloudflare.com/workers/configuration/cron-triggers/#remove-a-cron-trigger) |
| Configure renewable-energy execution locations | [Green Compute](https://developers.cloudflare.com/workers/configuration/cron-triggers/#green-compute) |
| Check trigger counts and execution budgets | [Workers limits](https://developers.cloudflare.com/workers/platform/limits/) |
| Schedule Workflow instances directly | [Schedule a Workflow directly](https://developers.cloudflare.com/workflows/build/trigger-workflows/#schedule-a-workflow-directly) |

Identify the target environment and the intended business timezone before choosing an expression. Review which schedules a deployment will replace, and use the documented propagation behavior when planning a rollout. For Green Compute, follow its account-level configuration rather than inferring settings from Worker placement.

See [api.md](./api.md) for implementation and [gotchas.md](./gotchas.md) for verification.

# Cron Triggers Troubleshooting

Investigate using the current documentation rather than copied limits or assumed delivery guarantees.

| Symptom or question | Documentation and check |
| --- | --- |
| Job runs at an unexpected time | Check [UTC execution](https://developers.cloudflare.com/workers/configuration/cron-triggers/#background) and [expression syntax](https://developers.cloudflare.com/workers/configuration/cron-triggers/#supported-cron-expressions); compare with the intended business timezone. |
| Schedule is missing after deployment | Check the handler, target environment, and propagation guidance in [Add a Cron Trigger](https://developers.cloudflare.com/workers/configuration/cron-triggers/#add-a-cron-trigger). |
| Removing or preserving schedules has an unexpected result | Review [Remove a Cron Trigger](https://developers.cloudflare.com/workers/configuration/cron-triggers/#remove-a-cron-trigger) before changing empty or omitted configuration. |
| Local invocation fails | Follow [Test Cron Triggers locally](https://developers.cloudflare.com/workers/configuration/cron-triggers/#test-cron-triggers-locally) for the supported endpoint, port, and query parameters. |
| Async work fails or completion status is surprising | Read [handler methods](https://developers.cloudflare.com/workers/runtime-apis/handlers/scheduled/#methods) and inspect [past events](https://developers.cloudflare.com/workers/configuration/cron-triggers/#view-past-events). |
| Job exceeds its execution budget | Check [Workers limits](https://developers.cloudflare.com/workers/platform/limits/) before choosing a smaller unit of work or [Workflows](../workflows/README.md). |
| Green Compute behavior differs from expectations | Read [Green Compute](https://developers.cloudflare.com/workers/configuration/cron-triggers/#green-compute) for its execution-location policy and account configuration. |

For repeated or partially completed business operations, decide how to identify work and recover safely before selecting storage. Test recovery after each side effect; a marker alone does not establish that the operation completed. Do not assume a particular automatic retry schedule or delivery guarantee without a documented contract.

See [patterns.md](./patterns.md) for coordination choices and [api.md](./api.md) for tests.

# Cron Triggers Patterns

Choose the execution boundary before writing a scheduled job; fetch the relevant integration docs for implementation.

| Need | Documentation |
| --- | --- |
| Periodic API sync, cleanup, reports, or health checks in a Worker | [Cron Triggers background](https://developers.cloudflare.com/workers/configuration/cron-triggers/#background) and [scheduled handler](https://developers.cloudflare.com/workers/runtime-apis/handlers/scheduled/) |
| Different jobs on different schedules | [Handle multiple cron triggers](https://developers.cloudflare.com/workers/runtime-apis/handlers/scheduled/#handle-multiple-cron-triggers) |
| Durable multi-step work started on a schedule | [Trigger Workflows](https://developers.cloudflare.com/workflows/build/trigger-workflows/) |
| Send work to a queue and implement its consumer | [Queues JavaScript APIs](https://developers.cloudflare.com/queues/configuration/javascript-apis/) |
| Coordinate state across invocations | [What are Durable Objects?](https://developers.cloudflare.com/durable-objects/concepts/what-are-durable-objects/) |
| Inspect whether scheduled work ran | [View past events](https://developers.cloudflare.com/workers/configuration/cron-triggers/#view-past-events) |

Keep the trigger separate from the business operation so manual recovery and scheduled execution can share it. Decide how partial progress is recorded, how repeated attempts affect side effects, and who owns completion reporting. When distributing a batch, distinguish successful enqueueing from successful processing.

Use [Workers limits](https://developers.cloudflare.com/workers/platform/limits/) to evaluate whether the job fits one invocation. If it needs durable steps, waiting, or explicit retry boundaries, inspect [Workflows](../workflows/README.md) before building those mechanisms in the handler.

See [api.md](./api.md) for tests and [gotchas.md](./gotchas.md) for operational checks.

# Cloudflare Cron Triggers

Use Cron Triggers to start periodic Worker jobs. Fetch the relevant current documentation before implementing; configuration, API signatures, examples, and limits belong in the docs.

- **Set up a recurring job:** [Cron Triggers](https://developers.cloudflare.com/workers/configuration/cron-triggers/) covers scheduling, deployment, and execution history.
- **Implement the job:** [Scheduled handler](https://developers.cloudflare.com/workers/runtime-apis/handlers/scheduled/) covers controller properties, asynchronous work, and multiple schedules.
- **Schedule durable work:** [Trigger Workflows](https://developers.cloudflare.com/workflows/build/trigger-workflows/) covers direct Workflow schedules and starting instances from a Worker. Check this before introducing a Worker whose only job is to start a Workflow.
- **Check capacity:** fetch [Workers limits](https://developers.cloudflare.com/workers/platform/limits/) for the target plan and invocation type.

## In This Reference

- [configuration.md](./configuration.md) — schedule setup, environments, removal, and Green Compute
- [api.md](./api.md) — handler implementation, asynchronous completion, and tests
- [patterns.md](./patterns.md) — choosing execution boundaries and integrations
- [gotchas.md](./gotchas.md) — investigating timing, failures, and repeated work

## See Also

- [Workflows](../workflows/README.md) — durable multi-step jobs
- [Queues](../queues/README.md) — asynchronous message processing

# D1 API Reference

Fetch the relevant API page before writing queries or assuming method signatures and return types.

| Task | Current documentation |
| --- | --- |
| Bind values and choose a query execution method | [Prepared statement methods](https://developers.cloudflare.com/d1/worker-api/prepared-statements/) |
| Execute batches and understand transaction rollback; use database sessions | [D1 Database API](https://developers.cloudflare.com/d1/worker-api/d1-database/) |
| Interpret results and query metadata | [Return objects](https://developers.cloudflare.com/d1/worker-api/return-object/) |
| Choose supported JavaScript values and TypeScript result types | [Workers Binding API](https://developers.cloudflare.com/d1/worker-api/) |
| Choose consistency constraints and carry bookmarks between requests | [Read replication and Sessions API](https://developers.cloudflare.com/d1/best-practices/read-replication/) |
| Query from a server-side script outside Workers | [REST query API](https://developers.cloudflare.com/api/resources/d1/subresources/database/methods/query/) |
| Handle query failures and transient errors | [Debug D1](https://developers.cloudflare.com/d1/observability/debug-d1/) and [retry queries](https://developers.cloudflare.com/d1/best-practices/retry-queries/) |
| Test database queries and apply migrations in tests | [Workers Vitest APIs: D1](https://developers.cloudflare.com/workers/testing/vitest-integration/test-apis/#d1) |

Bind untrusted values with prepared statements; do not interpolate them into SQL. Parameters do not replace identifiers: choose dynamic table, column, or sort names from an application-controlled allowlist.

D1 sessions provide sequential consistency for replicated queries. They are not a way to extend query execution limits. Choose the starting constraint or bookmark from the application's consistency requirements using the replication guide.

# D1 Configuration

Read the task's documentation before adding bindings or running database commands. Confirm the database and environment being targeted, particularly when applying migrations or importing data.

| Task | Current documentation |
| --- | --- |
| Create a database and attach a Worker binding | [Getting started](https://developers.cloudflare.com/d1/get-started/) |
| Configure binding fields and multiple databases | [Wrangler D1 configuration](https://developers.cloudflare.com/workers/wrangler/configuration/#d1-databases) |
| Separate staging and production databases | [D1 environments](https://developers.cloudflare.com/d1/configuration/environments/) |
| Create, track, and apply schema migrations | [Migrations](https://developers.cloudflare.com/d1/reference/migrations/) |
| Look up CLI flags for management, execution, and exports | [D1 Wrangler commands](https://developers.cloudflare.com/d1/wrangler-commands/) |
| Develop against local database state | [Local development](https://developers.cloudflare.com/d1/best-practices/local-development/) |
| Generate binding types | [Workers TypeScript](https://developers.cloudflare.com/workers/languages/typescript/) |
| Choose an ORM or query builder, including Drizzle | [D1 community projects](https://developers.cloudflare.com/d1/reference/community-projects/) (follow the integration's current setup guide) |
| Import or export SQL data | [Import and export data](https://developers.cloudflare.com/d1/best-practices/import-export-data/) |
| Enable replicas and use them through sessions | [Read replication](https://developers.cloudflare.com/d1/best-practices/read-replication/) |

Local migrations and data do not automatically update a remote database. Test against a separate staging database before a production migration. Naming another binding `DB_REPLICA` does not configure replica routing; follow the replication guide.

# D1 Gotchas & Troubleshooting

Use current documentation to diagnose the failure before changing query or database configuration.

| Symptom or question | What to check |
| --- | --- |
| Missing table, query exception, or constraint error | [Debug D1](https://developers.cloudflare.com/d1/observability/debug-d1/); verify the target binding, environment, and applied [migrations](https://developers.cloudflare.com/d1/reference/migrations/) |
| Boolean, date, or other binding type mismatch | [Workers Binding API type conversion](https://developers.cloudflare.com/d1/worker-api/) and [SQL support](https://developers.cloudflare.com/d1/sql-api/sql-statements/) |
| Foreign key failure during writes or migrations | [Foreign key enforcement and deferral](https://developers.cloudflare.com/d1/sql-api/foreign-keys/) |
| Slow queries, scans, or excessive rows read | [Indexes and query plans](https://developers.cloudflare.com/d1/best-practices/use-indexes/) and [metrics](https://developers.cloudflare.com/d1/observability/metrics-analytics/) |
| Query duration, statement, storage, or account limits | [Current limits](https://developers.cloudflare.com/d1/platform/limits/) |
| Unexpected usage charges or plan assumptions | [Pricing](https://developers.cloudflare.com/d1/platform/pricing/) |
| Stale reads after a write | [Sessions, bookmarks, and read replication](https://developers.cloudflare.com/d1/best-practices/read-replication/) |
| Transient query failures | [Retry guidance](https://developers.cloudflare.com/d1/best-practices/retry-queries/); check idempotency before retrying writes |
| Import/export failure or unsupported data | [Import/export behavior and limitations](https://developers.cloudflare.com/d1/best-practices/import-export-data/) |
| Local and deployed databases differ | [Local development](https://developers.cloudflare.com/d1/best-practices/local-development/) and [environment configuration](https://developers.cloudflare.com/d1/configuration/environments/) |

Continue to bind untrusted SQL values as described in [api.md](./api.md). Do not treat SQL injection as a recoverable database error or assume retries correct invalid queries.

# D1 Patterns & Best Practices

Use these guides to design the operation, then fetch [api.md](./api.md) for implementation references.

| Task | Current documentation |
| --- | --- |
| Design pagination, filters, joins, and aggregations | [Query a database](https://developers.cloudflare.com/d1/best-practices/query-d1/) and [supported SQL](https://developers.cloudflare.com/d1/sql-api/sql-statements/) |
| Reduce scans and inspect query plans | [Use indexes](https://developers.cloudflare.com/d1/best-practices/use-indexes/) |
| Batch writes or transform data | [Database API](https://developers.cloudflare.com/d1/worker-api/d1-database/) and [limits](https://developers.cloudflare.com/d1/platform/limits/) |
| Store and query event metadata | [Query JSON](https://developers.cloudflare.com/d1/sql-api/query-json/) |
| Evaluate a cache in front of D1 | [How KV works](https://developers.cloudflare.com/kv/concepts/how-kv-works/) |
| Choose shared or per-tenant databases | [D1 FAQs](https://developers.cloudflare.com/d1/reference/faq/) and [limits](https://developers.cloudflare.com/d1/platform/limits/) |
| Reduce read latency while preserving required consistency | [Read replication](https://developers.cloudflare.com/d1/best-practices/read-replication/) |
| Plan point-in-time recovery or portable backups | [Time Travel](https://developers.cloudflare.com/d1/reference/time-travel/) and [import/export](https://developers.cloudflare.com/d1/best-practices/import-export-data/) |

Keep result sets bounded and pagination ordering deterministic. Choose indexes from actual query plans. When splitting a large operation into batches, account for the loss of whole-operation atomicity across batches.

Authorize a tenant before selecting its database or rows; a request header alone is not proof of tenant membership. Application login sessions stored in tables are separate from D1's Sessions API.

Before caching reads, decide how stale data may be and how writes invalidate cached results. For replicated reads, choose session constraints and bookmark propagation based on read-after-write requirements rather than assuming every read sees the latest primary state.

# Cloudflare D1 Database

Use D1 for managed relational application data with SQLite semantics. For an existing external database, consider [Hyperdrive](../hyperdrive/); for per-entity coordination, consider [Durable Objects](https://developers.cloudflare.com/workers/platform/storage-options/#sql-in-durable-objects-vs-d1). See [storage options](https://developers.cloudflare.com/workers/platform/storage-options/) before choosing a product.

Read the relevant current documentation before implementing. These references route tasks to the source of truth rather than maintaining copies of APIs, configuration, or plan tables.

## Start here

- [Get started](https://developers.cloudflare.com/d1/get-started/): create a database, bind it to a Worker, and run a first query.
- [configuration.md](./configuration.md): bindings, environments, migrations, local development, and ORM integration.
- [api.md](./api.md): parameterized queries, batches, sessions, HTTP access, and testing.
- [patterns.md](./patterns.md): query design, caching, tenant isolation, replication, and recovery.
- [gotchas.md](./gotchas.md): errors, types, constraints, performance, and limits.

Check [limits](https://developers.cloudflare.com/d1/platform/limits/) and [pricing](https://developers.cloudflare.com/d1/platform/pricing/) for capacity, allowances, and plan availability; do not infer them from old examples.

# DDoS API

## Endpoints

### HTTP DDoS (L7)

```typescript
// Zone-level
PUT /zones/{zoneId}/rulesets/phases/ddos_l7/entrypoint
GET /zones/{zoneId}/rulesets/phases/ddos_l7/entrypoint

// Account-level (Enterprise Advanced)
PUT /accounts/{accountId}/rulesets/phases/ddos_l7/entrypoint
GET /accounts/{accountId}/rulesets/phases/ddos_l7/entrypoint
```

### Network DDoS (L3/4)

```typescript
// Account-level only
PUT /accounts/{accountId}/rulesets/phases/ddos_l4/entrypoint
GET /accounts/{accountId}/rulesets/phases/ddos_l4/entrypoint
```

## TypeScript SDK

**SDK Version**: Requires `cloudflare` >= 3.0.0 for ruleset phase methods.

```typescript
import Cloudflare from "cloudflare";

const client = new Cloudflare({ apiToken: process.env.CLOUDFLARE_API_TOKEN });

// STEP 1: Discover managed ruleset ID (required for overrides)
const allRulesets = await client.rulesets.list({ zone_id: zoneId });
const ddosRuleset = allRulesets.result.find(
  (r) => r.kind === "managed" && r.phase === "ddos_l7"
);
if (!ddosRuleset) throw new Error("DDoS managed ruleset not found");
const managedRulesetId = ddosRuleset.id;

// STEP 2: Get current HTTP DDoS configuration
const entrypointRuleset = await client.zones.rulesets.phases.entrypoint.get("ddos_l7", {
  zone_id: zoneId,
});

// STEP 3: Update HTTP DDoS ruleset with overrides
await client.zones.rulesets.phases.entrypoint.update("ddos_l7", {
  zone_id: zoneId,
  rules: [
    {
      action: "execute",
      expression: "true",
      action_parameters: {
        id: managedRulesetId, // From discovery step
        overrides: {
          sensitivity_level: "medium",
          action: "managed_challenge",
        },
      },
    },
  ],
});

// Network DDoS (account level, L3/4)
const l4Rulesets = await client.rulesets.list({ account_id: accountId });
const l4DdosRuleset = l4Rulesets.result.find(
  (r) => r.kind === "managed" && r.phase === "ddos_l4"
);
const l4Ruleset = await client.accounts.rulesets.phases.entrypoint.get("ddos_l4", {
  account_id: accountId,
});
```

## Alert Configuration

```typescript
interface DDoSAlertConfig {
  name: string;
  enabled: boolean;
  alert_type: "http_ddos_attack_alert" | "layer_3_4_ddos_attack_alert" 
    | "advanced_http_ddos_attack_alert" | "advanced_layer_3_4_ddos_attack_alert";
  filters?: {
    zones?: string[];
    hostnames?: string[];
    requests_per_second?: number;
    packets_per_second?: number;
    megabits_per_second?: number;
    ip_prefixes?: string[]; // CIDR
    ip_addresses?: string[];
    protocols?: string[];
  };
  mechanisms: {
    email?: Array<{ id: string }>;
    webhooks?: Array<{ id: string }>;
    pagerduty?: Array<{ id: string }>;
  };
}

// Create alert
await fetch(
  `https://api.cloudflare.com/client/v4/accounts/${accountId}/alerting/v3/policies`,
  {
    method: "POST",
    headers: {
      Authorization: `Bearer ${apiToken}`,
      "Content-Type": "application/json",
    },
    body: JSON.stringify(alertConfig),
  }
);
```

## Typed Override Examples

```typescript
// Override by category
interface CategoryOverride {
  action: "execute";
  expression: string;
  action_parameters: {
    id: string;
    overrides: {
      categories?: Array<{
        category: "http-flood" | "http-anomaly" | "udp-flood" | "syn-flood";
        sensitivity_level?: "default" | "medium" | "low" | "eoff";
        action?: "block" | "managed_challenge" | "challenge" | "log";
      }>;
    };
  };
}

// Override by rule ID
interface RuleOverride {
  action: "execute";
  expression: string;
  action_parameters: {
    id: string;
    overrides: {
      rules?: Array<{
        id: string;
        action?: "block" | "managed_challenge" | "challenge" | "log";
        sensitivity_level?: "default" | "medium" | "low" | "eoff";
      }>;
    };
  };
}

// Example: Override specific adaptive rule
const adaptiveOverride: RuleOverride = {
  action: "execute",
  expression: "true",
  action_parameters: {
    id: managedRulesetId,
    overrides: {
      rules: [
        { id: "...adaptive-origins-rule-id...", sensitivity_level: "low" },
      ],
    },
  },
};
```

See [patterns.md](./patterns.md) for complete implementation patterns.

# DDoS Configuration

## Dashboard Setup

1. Navigate to Security > DDoS
2. Select HTTP DDoS or Network-layer DDoS
3. Configure sensitivity & action per ruleset/category/rule
4. Apply overrides with optional expressions (Enterprise Advanced)
5. Enable Adaptive DDoS toggle (Enterprise/Enterprise Advanced, requires 7 days traffic history)

## Rule Structure

```typescript
interface DDoSOverride {
  description: string;
  rules: Array<{
    action: "execute";
    expression: string; // Custom expression (Enterprise Advanced) or "true" for all
    action_parameters: {
      id: string; // Managed ruleset ID (discover via api.md)
      overrides: {
        sensitivity_level?: "default" | "medium" | "low" | "eoff";
        action?: "block" | "managed_challenge" | "challenge" | "log"; // log = Enterprise Advanced only
        categories?: Array<{
          category: string; // e.g., "http-flood", "udp-flood"
          sensitivity_level?: string;
        }>;
        rules?: Array<{
          id: string;
          action?: string;
          sensitivity_level?: string;
        }>;
      };
    };
  }>;
}
```

## Expression Availability

| Plan | Custom Expressions | Example |
|------|-------------------|---------|
| Free/Pro/Business | ✗ | Use `"true"` only |
| Enterprise | ✗ | Use `"true"` only |
| Enterprise Advanced | ✓ | `ip.src in {...}`, `http.request.uri.path matches "..."` |

## Sensitivity Mapping

| UI | API | Threshold |
|----|-----|-----------|
| High | `default` | Most aggressive |
| Medium | `medium` | Balanced |
| Low | `low` | Less aggressive |
| Essentially Off | `eoff` | Minimal mitigation |

## Common Categories

- `http-flood`, `http-anomaly` (L7)
- `udp-flood`, `syn-flood`, `dns-flood` (L3/4)

## Override Precedence

Multiple override layers apply in this order (higher precedence wins):

```
Zone-level > Account-level
Individual Rule > Category > Global sensitivity/action
```

**Example**: Zone rule for `/api/*` overrides account-level global settings.

## Adaptive DDoS Profiles

**Availability**: Enterprise, Enterprise Advanced  
**Learning period**: 7 days of traffic history required

| Profile Type | Description | Detects |
|--------------|-------------|---------|
| **Origins** | Traffic patterns per origin server | Anomalous requests to specific origins |
| **User-Agents** | Traffic patterns per User-Agent | Malicious/anomalous user agent strings |
| **Locations** | Traffic patterns per geo-location | Attacks from specific countries/regions |
| **Protocols** | Traffic patterns per protocol (L3/4) | Protocol-specific flood attacks |

Configure by targeting specific adaptive rule IDs via API (see api.md#typed-override-examples).

## Alerting

Configure via Notifications:
- Alert types: `http_ddos_attack_alert`, `layer_3_4_ddos_attack_alert`, `advanced_*` variants
- Filters: zones, hostnames, RPS/PPS/Mbps thresholds, IPs, protocols
- Mechanisms: email, webhooks, PagerDuty

See [api.md](./api.md#alert-configuration) for API examples.

# DDoS Gotchas

## Common Errors

### "False positives blocking legitimate traffic"

**Cause**: Sensitivity too high, wrong action, or missing exceptions  
**Solution**:
1. Lower sensitivity for specific rule/category
2. Use `log` action first to validate (Enterprise Advanced)
3. Add exception with custom expression (e.g., allowlist IPs)
4. Query flagged requests via GraphQL Analytics API to identify patterns

### "Attacks getting through"

**Cause**: Sensitivity too low or wrong action  
**Solution**: Increase to `default` sensitivity and use `block` action:
```typescript
const config = {
  rules: [{
    expression: "true",
    action: "execute",
    action_parameters: { id: managedRulesetId, overrides: { sensitivity_level: "default", action: "block" } },
  }],
};
```

### "Adaptive rules not working"

**Cause**: Insufficient traffic history (needs 7 days)  
**Solution**: Wait for baseline to establish, check dashboard for adaptive rule status

### "Zone override ignored"

**Cause**: Account overrides conflict with zone overrides  
**Solution**: Configure at zone level OR remove zone overrides to use account-level

### "Log action not available"

**Cause**: Not on Enterprise Advanced DDoS plan  
**Solution**: Use `managed_challenge` with low sensitivity for testing

### "Rule limit exceeded"

**Cause**: Too many override rules (Free/Pro/Business: 1, Enterprise Advanced: 10)  
**Solution**: Combine conditions in single expression using `and`/`or`

### "Cannot override rule"

**Cause**: Rule is read-only  
**Solution**: Check API response for read-only indicator, use different rule

### "Cannot disable DDoS protection"

**Cause**: DDoS managed rulesets cannot be fully disabled (always-on protection)  
**Solution**: Set `sensitivity_level: "eoff"` for minimal mitigation

### "Expression not allowed"

**Cause**: Custom expressions require Enterprise Advanced plan  
**Solution**: Use `expression: "true"` for all traffic, or upgrade plan

### "Managed ruleset not found"

**Cause**: Zone/account doesn't have DDoS managed ruleset, or incorrect phase  
**Solution**: Verify ruleset exists via `client.rulesets.list()`, check phase name (`ddos_l7` or `ddos_l4`)

## API Error Codes

| Error Code | Message | Cause | Solution |
|------------|---------|-------|----------|
| 10000 | Authentication error | Invalid/missing API token | Check token has DDoS permissions |
| 81000 | Ruleset validation failed | Invalid rule structure | Verify `action_parameters.id` is managed ruleset ID |
| 81020 | Expression not allowed | Custom expressions on wrong plan | Use `"true"` or upgrade to Enterprise Advanced |
| 81021 | Rule limit exceeded | Too many override rules | Reduce rules or upgrade (Enterprise Advanced: 10) |
| 81022 | Invalid sensitivity level | Wrong sensitivity value | Use: `default`, `medium`, `low`, `eoff` |
| 81023 | Invalid action | Wrong action for plan | Enterprise Advanced only: `log` action |

## Limits

| Resource/Limit | Free/Pro/Business | Enterprise | Enterprise Advanced |
|----------------|-------------------|------------|---------------------|
| Override rules per zone | 1 | 1 | 10 |
| Custom expressions | ✗ | ✗ | ✓ |
| Log action | ✗ | ✗ | ✓ |
| Adaptive DDoS | ✗ | ✓ | ✓ |
| Traffic history required | - | 7 days | 7 days |

## Tuning Strategy

1. Start with `log` action + `medium` sensitivity
2. Monitor for 24-48 hours
3. Identify false positives, add exceptions
4. Gradually increase to `default` sensitivity
5. Change action from `log` → `managed_challenge` → `block`
6. Document all adjustments

## Best Practices

- Test during low-traffic periods
- Use zone-level for per-site tuning
- Reference IP lists for easier management
- Set appropriate alert thresholds (avoid noise)
- Combine with WAF for layered defense
- Avoid over-tuning (keep config simple)

See [patterns.md](./patterns.md) for progressive rollout examples.

# DDoS Protection Patterns

## Allowlist Trusted IPs

```typescript
const config = {
  description: "Allowlist trusted IPs",
  rules: [{
    expression: "ip.src in { 203.0.113.0/24 192.0.2.1 }",
    action: "execute",
    action_parameters: {
      id: managedRulesetId,
      overrides: { sensitivity_level: "eoff" },
    },
  }],
};

await client.accounts.rulesets.phases.entrypoint.update("ddos_l7", {
  account_id: accountId,
  ...config,
});
```

## Route-specific Sensitivity

```typescript
const config = {
  description: "Route-specific protection",
  rules: [
    {
      expression: "not http.request.uri.path matches \"^/api/\"",
      action: "execute",
      action_parameters: {
        id: managedRulesetId,
        overrides: { sensitivity_level: "default", action: "block" },
      },
    },
    {
      expression: "http.request.uri.path matches \"^/api/\"",
      action: "execute",
      action_parameters: {
        id: managedRulesetId,
        overrides: { sensitivity_level: "low", action: "managed_challenge" },
      },
    },
  ],
};
```

## Progressive Enhancement

```typescript
enum ProtectionLevel { MONITORING = "monitoring", LOW = "low", MEDIUM = "medium", HIGH = "high" }

const levelConfig = {
  [ProtectionLevel.MONITORING]: { action: "log", sensitivity: "eoff" },
  [ProtectionLevel.LOW]: { action: "managed_challenge", sensitivity: "low" },
  [ProtectionLevel.MEDIUM]: { action: "managed_challenge", sensitivity: "medium" },
  [ProtectionLevel.HIGH]: { action: "block", sensitivity: "default" },
} as const;

async function setProtectionLevel(zoneId: string, level: ProtectionLevel, rulesetId: string, client: Cloudflare) {
  const settings = levelConfig[level];
  return client.zones.rulesets.phases.entrypoint.update("ddos_l7", {
    zone_id: zoneId,
    rules: [{
      expression: "true",
      action: "execute",
      action_parameters: { id: rulesetId, overrides: { action: settings.action, sensitivity_level: settings.sensitivity } },
    }],
  });
}
```

## Dynamic Response to Attacks

```typescript
interface Env { CLOUDFLARE_API_TOKEN: string; ZONE_ID: string; KV: KVNamespace; }

export default {
  async fetch(request: Request, env: Env): Promise<Response> {
    if (request.url.includes("/attack-detected")) {
      const attackData = await request.json();
      await env.KV.put(`attack:${Date.now()}`, JSON.stringify(attackData), { expirationTtl: 86400 });
      const recentAttacks = await getRecentAttacks(env.KV);
      if (recentAttacks.length > 5) {
        await setProtectionLevel(env.ZONE_ID, ProtectionLevel.HIGH, managedRulesetId, client);
        return new Response("Protection increased");
      }
    }
    return new Response("OK");
  },
  async scheduled(event: ScheduledEvent, env: Env): Promise<void> {
    const recentAttacks = await getRecentAttacks(env.KV);
    if (recentAttacks.length === 0) await setProtectionLevel(env.ZONE_ID, ProtectionLevel.MEDIUM, managedRulesetId, client);
  },
};
```

## Multi-rule Tiered Protection (Enterprise Advanced)

```typescript
const config = {
  description: "Multi-tier DDoS protection",
  rules: [
    {
      expression: "not ip.src in $known_ips and not cf.bot_management.score gt 30",
      action: "execute",
      action_parameters: { id: managedRulesetId, overrides: { sensitivity_level: "default", action: "block" } },
    },
    {
      expression: "cf.bot_management.verified_bot",
      action: "execute",
      action_parameters: { id: managedRulesetId, overrides: { sensitivity_level: "medium", action: "managed_challenge" } },
    },
    {
      expression: "ip.src in $trusted_ips",
      action: "execute",
      action_parameters: { id: managedRulesetId, overrides: { sensitivity_level: "low" } },
    },
  ],
};
```

## Defense in Depth

Layered security stack: DDoS + WAF + Rate Limiting + Bot Management.

```typescript
// Layer 1: DDoS (volumetric attacks)
await client.zones.rulesets.phases.entrypoint.update("ddos_l7", {
  zone_id: zoneId,
  rules: [{ expression: "true", action: "execute", action_parameters: { id: ddosRulesetId, overrides: { sensitivity_level: "medium" } } }],
});

// Layer 2: WAF (exploit protection)
await client.zones.rulesets.phases.entrypoint.update("http_request_firewall_managed", {
  zone_id: zoneId,
  rules: [{ expression: "true", action: "execute", action_parameters: { id: wafRulesetId } }],
});

// Layer 3: Rate Limiting (abuse prevention)
await client.zones.rulesets.phases.entrypoint.update("http_ratelimit", {
  zone_id: zoneId,
  rules: [{ expression: "http.request.uri.path eq \"/api/login\"", action: "block", ratelimit: { characteristics: ["ip.src"], period: 60, requests_per_period: 5 } }],
});

// Layer 4: Bot Management (automation detection)
await client.zones.rulesets.phases.entrypoint.update("http_request_sbfm", {
  zone_id: zoneId,
  rules: [{ expression: "cf.bot_management.score lt 30", action: "managed_challenge" }],
});
```

## Cache Strategy for DDoS Mitigation

Exclude query strings from cache key to counter randomized query parameter attacks.

```typescript
const cacheRule = {
  expression: "http.request.uri.path matches \"^/api/\"",
  action: "set_cache_settings",
  action_parameters: {
    cache: true,
    cache_key: { ignore_query_strings_order: true, custom_key: { query_string: { exclude: { all: true } } } },
  },
};

await client.zones.rulesets.phases.entrypoint.update("http_request_cache_settings", { zone_id: zoneId, rules: [cacheRule] });
```

**Rationale**: Attackers randomize query strings (`?random=123456`) to bypass cache. Excluding query params ensures cache hits absorb attack traffic.

See [configuration.md](./configuration.md) for rule structure details.

# Cloudflare DDoS Protection

Autonomous, always-on protection against DDoS attacks across L3/4 and L7.

## Protection Types

- **HTTP DDoS (L7)**: Protects HTTP/HTTPS traffic, phase `ddos_l7`, zone/account level
- **Network DDoS (L3/4)**: UDP/SYN/DNS floods, phase `ddos_l4`, account level only
- **Adaptive DDoS**: Learns 7-day baseline, detects deviations, 4 profile types (Origins, User-Agents, Locations, Protocols)

## Plan Availability

| Feature | Free | Pro | Business | Enterprise | Enterprise Advanced |
|---------|------|-----|----------|------------|---------------------|
| HTTP DDoS (L7) | ✓ | ✓ | ✓ | ✓ | ✓ |
| Network DDoS (L3/4) | ✓ | ✓ | ✓ | ✓ | ✓ |
| Override rules | 1 | 1 | 1 | 1 | 10 |
| Custom expressions | ✗ | ✗ | ✗ | ✗ | ✓ |
| Log action | ✗ | ✗ | ✗ | ✗ | ✓ |
| Adaptive DDoS | ✗ | ✗ | ✗ | ✓ | ✓ |
| Alert filters | Basic | Basic | Basic | Advanced | Advanced |

## Actions & Sensitivity

- **Actions**: `block`, `managed_challenge`, `challenge`, `log` (Enterprise Advanced only)
- **Sensitivity**: `default` (high), `medium`, `low`, `eoff` (essentially off)
- **Override**: By category/tag or individual rule ID
- **Scope**: Zone-level overrides take precedence over account-level

## Reading Order

| File | Purpose | Start Here If... |
|------|---------|------------------|
| [configuration.md](./configuration.md) | Dashboard setup, rule structure, adaptive profiles | You're setting up DDoS protection for the first time |
| [api.md](./api.md) | API endpoints, SDK usage, ruleset ID discovery | You're automating configuration or need programmatic access |
| [patterns.md](./patterns.md) | Protection strategies, defense-in-depth, dynamic response | You need implementation patterns or layered security |
| [gotchas.md](./gotchas.md) | False positives, tuning, error handling | You're troubleshooting or optimizing existing protection |

## See Also
- [waf](../waf/) - Application-layer security rules
- [bot-management](../bot-management/) - Bot detection and mitigation

# DO Storage API

Check the class’s backend before choosing operations; storage APIs and recovery capabilities differ.

Fetch the relevant current documentation before implementing or reviewing changes.

| Task | Documentation |
|------|---------------|
| Use SQL cursors, bound parameters, supported SQL, or database size | [SQLite storage API](https://developers.cloudflare.com/durable-objects/api/sqlite-storage-api/) |
| Use synchronous or asynchronous key-value methods on SQLite | [SQLite storage API](https://developers.cloudflare.com/durable-objects/api/sqlite-storage-api/) |
| Maintain asynchronous KV operations on a legacy backend | [Legacy KV storage API](https://developers.cloudflare.com/durable-objects/api/legacy-kv-storage-api/) |
| Review transactions, write coalescing, storage options, or cleanup | [SQLite storage API](https://developers.cloudflare.com/durable-objects/api/sqlite-storage-api/); [Legacy KV storage API](https://developers.cloudflare.com/durable-objects/api/legacy-kv-storage-api/) |
| Create bookmarks or restore SQLite data | [SQLite storage API](https://developers.cloudflare.com/durable-objects/api/sqlite-storage-api/) |
| Schedule, inspect, or cancel an alarm | [Alarms](https://developers.cloudflare.com/durable-objects/api/alarms/) |

# DO Storage Configuration

Prefer SQLite for new classes. Inspect an existing class’s backend and lifecycle configuration before changing either.

Fetch the relevant current documentation before implementing or reviewing changes.

| Task | Documentation |
|------|---------------|
| Create a SQLite-backed class, binding, and generated types | [Getting started](https://developers.cloudflare.com/durable-objects/get-started/) |
| Choose storage and manage class exports | [Class exports](https://developers.cloudflare.com/durable-objects/reference/durable-objects-migrations/) |
| Maintain legacy migration configuration | [Legacy class migrations](https://developers.cloudflare.com/durable-objects/reference/durable-object-class-migrations-legacy/) |
| Initialize schemas or evolve application tables | [Rules of Durable Objects](https://developers.cloudflare.com/durable-objects/best-practices/rules-of-durable-objects/); [SQLite storage API](https://developers.cloudflare.com/durable-objects/api/sqlite-storage-api/) |
| Set placement hints or jurisdiction constraints | [Data location](https://developers.cloudflare.com/durable-objects/reference/data-location/) |
| Configure CPU allowances and check storage constraints | [Limits](https://developers.cloudflare.com/durable-objects/platform/limits/) |

A class configuration change is not an application-data migration. Check the documented backend transition constraints in [Class exports](https://developers.cloudflare.com/durable-objects/reference/durable-objects-migrations/) before planning a backend change.

# DO Storage Troubleshooting

Identify the backend and failing operation before applying concurrency or recovery guidance.

Fetch the relevant current documentation before implementing or reviewing changes.

| Task | Documentation |
|------|---------------|
| Input/output gates, write coalescing, external I/O races, or storage options | [Rules of Durable Objects](https://developers.cloudflare.com/durable-objects/best-practices/rules-of-durable-objects/); [SQLite storage API](https://developers.cloudflare.com/durable-objects/api/sqlite-storage-api/); [Legacy KV storage API](https://developers.cloudflare.com/durable-objects/api/legacy-kv-storage-api/) |
| SQL transactions, synchronous callbacks, parameter types, or numeric precision | [SQLite storage API](https://developers.cloudflare.com/durable-objects/api/sqlite-storage-api/) |
| Alarm cancellation and storage deletion | [Alarms](https://developers.cloudflare.com/durable-objects/api/alarms/); [SQLite storage API](https://developers.cloudflare.com/durable-objects/api/sqlite-storage-api/) |
| Slow queries, indexing, caching, or initialization | [Rules of Durable Objects](https://developers.cloudflare.com/durable-objects/best-practices/rules-of-durable-objects/); [Durable Object State](https://developers.cloudflare.com/durable-objects/api/state/) |
| Storage limits or CPU exhaustion | [Limits](https://developers.cloudflare.com/durable-objects/platform/limits/) |
| Storage charges and operation accounting | [Pricing](https://developers.cloudflare.com/durable-objects/platform/pricing/) |
| Overload, storage timeouts, or object resets | [Troubleshooting](https://developers.cloudflare.com/durable-objects/observability/troubleshooting/); [Error handling](https://developers.cloudflare.com/durable-objects/best-practices/error-handling/) |

# DO Storage Patterns

Persist essential state and treat memory as a reconstructible cache. Coordinate related updates within the storage and concurrency guarantees of the selected backend.

Fetch the relevant current documentation before implementing or reviewing changes.

| Task | Documentation |
|------|---------------|
| Schema initialization, migrations, indexes, caching, or parent-child coordination | [Rules of Durable Objects](https://developers.cloudflare.com/durable-objects/best-practices/rules-of-durable-objects/) |
| Counters, transactions, and atomic updates | [SQLite storage API](https://developers.cloudflare.com/durable-objects/api/sqlite-storage-api/); [Legacy KV storage API](https://developers.cloudflare.com/durable-objects/api/legacy-kv-storage-api/); [Counter example](https://developers.cloudflare.com/durable-objects/examples/build-a-counter/) |
| Batch processing or multiple scheduled events | [Alarms](https://developers.cloudflare.com/durable-objects/api/alarms/); [Batching example](https://developers.cloudflare.com/durable-objects/examples/alarms-api/) |
| Cleanup and expiration | [Time to Live example](https://developers.cloudflare.com/durable-objects/examples/durable-object-ttl/); [SQLite storage API](https://developers.cloudflare.com/durable-objects/api/sqlite-storage-api/) |
| Design application-specific rate limiting | [Rules of Durable Objects](https://developers.cloudflare.com/durable-objects/best-practices/rules-of-durable-objects/); [SQLite storage API](https://developers.cloudflare.com/durable-objects/api/sqlite-storage-api/) |

Verify persistence, isolation, and rollback behavior with the [testing guidance](testing.md).

# Cloudflare Durable Objects Storage

Use SQLite for new classes. Existing KV-backed classes need their matching API reference; using key-value methods does not by itself identify the backend.

Fetch the relevant current documentation before implementing or reviewing changes.

| Task | Documentation |
|------|---------------|
| Choose SQL, key-value access, transactions, or recovery APIs | [SQLite storage API](https://developers.cloudflare.com/durable-objects/api/sqlite-storage-api/); [Legacy KV storage API](https://developers.cloudflare.com/durable-objects/api/legacy-kv-storage-api/) |
| Configure the backend, class lifecycle, and placement | [Configuration](configuration.md) |
| Find operation semantics and storage options | [API routing](api.md) |
| Design schemas, caches, scheduled work, or cleanup | [Patterns](patterns.md) |
| Diagnose concurrency, limits, and billing | [Troubleshooting](gotchas.md) |
| Verify storage behavior in the Workers runtime | [Testing](testing.md) |

For object routing, WebSockets, and coordination design, see the [Durable Objects skill](../../../durable-objects/SKILL.md).

# DO Storage Testing

Choose tests around persistence, rollback, instance isolation, and scheduled-work behavior. Inspect installed test packages and configuration before changing the suite.

Fetch the relevant current documentation before implementing or reviewing changes.

| Task | Documentation |
|------|---------------|
| Set up or migrate a test suite, choose helpers, and manage isolation | [Testing Durable Objects](../../../durable-objects/references/testing.md) |
| Exercise RPC, SQLite storage, and alarms | [Testing Durable Objects example](https://developers.cloudflare.com/durable-objects/examples/testing-with-durable-objects/) |
| Determine the storage or recovery contract to verify | [SQLite storage API](https://developers.cloudflare.com/durable-objects/api/sqlite-storage-api/); [Legacy KV storage API](https://developers.cloudflare.com/durable-objects/api/legacy-kv-storage-api/) |

Use the current test documentation for helper signatures and runtime limitations. For point-in-time recovery tests, check both the storage API and the test runtime’s supported behavior before assuming a restart reproduces production recovery.

# Email Routing APIs

Fetch the relevant API page before writing code; do not infer sending types or recipient restrictions from incoming-mail APIs.

| Task | Documentation |
| --- | --- |
| Manage routing settings, rules, and destination addresses programmatically | [Email Routing REST API](https://developers.cloudflare.com/api/resources/email_routing/) |
| Read incoming message metadata; forward, reply, or reject | [Email handler API](https://developers.cloudflare.com/email-service/api/route-emails/email-handler/) |
| Send from a Worker, including attachments or existing raw MIME | [Sending Workers API](https://developers.cloudflare.com/email-service/api/send-emails/workers-api/) |
| Restrict a sending binding's senders or recipients | [Configure send bindings](https://developers.cloudflare.com/email-service/configuration/send-bindings/) |
| Send from an external application | [Sending REST API](https://developers.cloudflare.com/email-service/api/send-emails/rest-api/) or [SMTP](https://developers.cloudflare.com/email-service/api/send-emails/smtp/) |

For incoming messages, distinguish SMTP envelope addresses from message headers. Use [Email Workers API guidance](../email-workers/api.md) for processing and [authentication docs](https://developers.cloudflare.com/email-service/concepts/email-authentication/) for identity checks.

# Email Routing Setup

Fetch the setup page matching the operation. Email Sending and Email Routing have separate domain configuration; enabling one is not a substitute for configuring the other.

| Task | Documentation |
| --- | --- |
| Onboard a routing domain and deploy/connect an Email Worker | [Route emails](https://developers.cloudflare.com/email-service/get-started/route-emails/) |
| Inspect DNS records, conflicts, verification, or disable routing | [Domain configuration](https://developers.cloudflare.com/email-service/configuration/domains/) |
| Verify forwarding destinations and manage routing or catch-all rules | [Routing rules and addresses](https://developers.cloudflare.com/email-service/configuration/email-routing-addresses/) |
| Configure a subdomain | [Subdomains](https://developers.cloudflare.com/email-service/configuration/subdomains/) |
| Configure outbound email | [Send emails](https://developers.cloudflare.com/email-service/get-started/send-emails/) and [send binding restrictions](https://developers.cloudflare.com/email-service/configuration/send-bindings/) |
| Test an incoming email locally | [Local routing development](https://developers.cloudflare.com/email-service/local-development/routing/) |
| Add Worker storage, types, secrets, or environments | [Email Workers configuration](../email-workers/configuration.md) |

# Email Routing Troubleshooting

Start with the message's activity log to distinguish routing, authentication, and delivery failures, then fetch the matching documentation.

| Symptom or question | Documentation |
| --- | --- |
| Rule disabled, wrong destination, or catch-all behavior | [Routing rules and verified addresses](https://developers.cloudflare.com/email-service/configuration/email-routing-addresses/) |
| DNS conflict or domain not configured | [Domain configuration](https://developers.cloudflare.com/email-service/configuration/domains/) |
| SPF, DKIM, or DMARC failure | [Authentication troubleshooting](https://developers.cloudflare.com/email-service/reference/troubleshooting/) |
| Message missing, rejected, dropped, or delivery failed | [Email logs](https://developers.cloudflare.com/email-service/observability/logs/) |
| Quotas, message sizes, routing capacity, or Worker resource exhaustion | [Current limits](https://developers.cloudflare.com/email-service/platform/limits/) |
| Sending costs and verified-destination allowances | [Pricing](https://developers.cloudflare.com/email-service/platform/pricing/) |
| Stream, parser, reply, or Worker execution error | [Email Workers troubleshooting](../email-workers/gotchas.md) |

Do not use a sender-address string as proof of authentication. Inspect the authentication results described in the logs and authentication docs.

# Email Routing Patterns

Prefer a routing rule when the destination depends only on the email address. Use an Email Worker for decisions based on message content or application state.

| Task | Documentation |
| --- | --- |
| Address-based forwarding, catch-all, or subaddressing | [Routing rules and addresses](https://developers.cloudflare.com/email-service/configuration/email-routing-addresses/) |
| Recipient/subject routing, multiple destinations, rejection, or automatic replies | [Email handler actions](https://developers.cloudflare.com/email-service/api/route-emails/email-handler/) |
| Filter unwanted messages | [Spam filtering](https://developers.cloudflare.com/email-service/examples/email-routing/spam-filtering/) |
| Parse MIME, extract attachments, archive mail, or notify an application | [Email Workers patterns](../email-workers/patterns.md) |
| Send outbound attachments | [Email attachments](https://developers.cloudflare.com/email-service/examples/email-sending/email-attachments/) |

Verify all forwarding destinations. For delayed responses after processing or human review, use the outbound sending API; an incoming event's reply operation belongs to that event.

# Email Routing

Use routing rules for address-based forwarding; use an Email Worker when incoming mail needs custom processing. Fetch the linked docs before implementing APIs, DNS, configuration, or limits.

| Task | Start here |
| --- | --- |
| Forward incoming mail to an existing mailbox | [Route emails](https://developers.cloudflare.com/email-service/get-started/route-emails/) |
| Manage addresses, verification, catch-all rules, or subaddressing | [Routing rules and addresses](https://developers.cloudflare.com/email-service/configuration/email-routing-addresses/) |
| Filter, parse, reply to, or store incoming mail | [Email Workers](../email-workers/README.md) |
| Send a new outbound message | [Send emails](https://developers.cloudflare.com/email-service/get-started/send-emails/) — Workers binding, REST API, or SMTP |

Forwarding requires verified destinations. Replying within an incoming email event and sending a new outbound message have different requirements; use the relevant API docs.

## Reference map

- [Configuration](configuration.md): domains, rules, deployment, and local testing.
- [API](api.md): routing management and inbound/outbound operations.
- [Patterns](patterns.md): filtering, parsing, storage, and notifications.
- [Troubleshooting](gotchas.md): authentication, delivery, and current limits.

# Email Workers APIs

Fetch the relevant page for current interfaces and return types.

| Task | Documentation |
| --- | --- |
| Implement the handler; inspect envelope addresses, headers, raw content, or size | [Email handler API](https://developers.cloudflare.com/email-service/api/route-emails/email-handler/) |
| Forward, add forwarding headers, reject, or reply with MIME and threading | [Email actions and reply requirements](https://developers.cloudflare.com/email-service/api/route-emails/email-handler/) |
| Parse MIME bodies and attachments | [Email handler parsing guidance](https://developers.cloudflare.com/email-service/api/route-emails/email-handler/) — follow its postal-mime reference |
| Send new outbound mail or an existing raw MIME message | [Sending Workers API](https://developers.cloudflare.com/email-service/api/send-emails/workers-api/) |
| Configure sender and recipient restrictions | [Send bindings](https://developers.cloudflare.com/email-service/configuration/send-bindings/) |
| Set outbound headers | [Email headers](https://developers.cloudflare.com/email-service/reference/headers/) |
| Generate Worker and binding types | [Workers TypeScript](https://developers.cloudflare.com/workers/languages/typescript/) |

Envelope addresses describe SMTP transport; message headers describe the message. Neither an address comparison nor a display header replaces [email authentication](https://developers.cloudflare.com/email-service/concepts/email-authentication/).

# Email Workers Configuration

An incoming routing rule connects an address to a Worker. Add an outbound sending binding when the application needs the sending API.

| Task | Documentation |
| --- | --- |
| Create, deploy, and connect an email-processing Worker | [Route emails](https://developers.cloudflare.com/email-service/get-started/route-emails/) |
| Verify destinations, configure rules, or check DNS | [Email Routing configuration](../email-routing/configuration.md) |
| Configure outbound sending and address restrictions | [Sending Workers API](https://developers.cloudflare.com/email-service/api/send-emails/workers-api/) and [send bindings](https://developers.cloudflare.com/email-service/configuration/send-bindings/) |
| Simulate incoming messages | [Local routing development](https://developers.cloudflare.com/email-service/local-development/routing/) |
| Test outbound messages and attachment behavior | [Local sending development](https://developers.cloudflare.com/email-service/local-development/sending/) |
| Configure KV, R2, D1, variables, or environments | [Wrangler configuration](https://developers.cloudflare.com/workers/wrangler/configuration/) |
| Generate runtime and binding types | [Workers TypeScript](https://developers.cloudflare.com/workers/languages/typescript/) |
| Store credentials | [Workers secrets](https://developers.cloudflare.com/workers/configuration/secrets/) |

Follow the handler docs for MIME library requirements. Local sending simulation and remote sending have different effects: remote bindings deliver real email.

# Email Workers Troubleshooting

| Symptom or decision | Documentation |
| --- | --- |
| Raw stream already consumed or locked | [ReadableStream](https://developers.cloudflare.com/workers/runtime-apis/streams/readablestream/) and [handler parsing guidance](https://developers.cloudflare.com/email-service/api/route-emails/email-handler/) |
| Forwarding or reply exception; unsupported forwarding headers | [Email handler actions and requirements](https://developers.cloudflare.com/email-service/api/route-emails/email-handler/) |
| Unverified destination or disabled rule | [Routing rules and addresses](https://developers.cloudflare.com/email-service/configuration/email-routing-addresses/) |
| Sender identity or authentication failure | [Email authentication](https://developers.cloudflare.com/email-service/concepts/email-authentication/) and [troubleshooting](https://developers.cloudflare.com/email-service/reference/troubleshooting/) |
| Sending validation, attachment, or recipient error | [Sending API errors](https://developers.cloudflare.com/email-service/api/send-emails/workers-api/) |
| Local test or binary attachment issue | [Local routing](https://developers.cloudflare.com/email-service/local-development/routing/) and [local sending](https://developers.cloudflare.com/email-service/local-development/sending/) |
| CPU, memory, message-size, or reply limits | [Email limits](https://developers.cloudflare.com/email-service/platform/limits/) and [Workers limits](https://developers.cloudflare.com/workers/platform/limits/) |
| Background work or unhandled error | [Execution context](https://developers.cloudflare.com/workers/runtime-apis/context/) and [Workers logs](https://developers.cloudflare.com/workers/observability/logs/) |
| Mail accepted but missing at the destination | [Email activity logs](https://developers.cloudflare.com/email-service/observability/logs/) |

Raw content is single-use: reuse buffered content if multiple operations need it, and account for memory limits. `waitUntil()` extends execution lifetime; it does not remove CPU or memory limits. Diagnose reply failures against the incoming message's requirements, not just the sending domain's DNS.

# Email Workers Patterns

Fetch the workflow page, then adapt it to the application's routing and storage requirements.

| Task | Documentation |
| --- | --- |
| Route by recipient or subject, forward to multiple destinations, or reject | [Email handler actions](https://developers.cloudflare.com/email-service/api/route-emails/email-handler/) |
| Parse MIME bodies and attachments | [Email handler parsing guidance](https://developers.cloudflare.com/email-service/api/route-emails/email-handler/) |
| Filter incoming mail | [Spam filtering](https://developers.cloudflare.com/email-service/examples/email-routing/spam-filtering/) |
| Reply within the incoming event with threading | [Reply requirements and examples](https://developers.cloudflare.com/email-service/api/route-emails/email-handler/) |
| Archive metadata in KV or enqueue mail for later processing | [Email storage and processing](https://developers.cloudflare.com/email-service/examples/email-routing/email-storage/) |
| Store raw mail or extracted attachment bytes in R2 | [R2 Workers API](https://developers.cloudflare.com/r2/api/workers/workers-api-reference/) |
| Notify a webhook or schedule work within the invocation lifetime | [Fetch](https://developers.cloudflare.com/workers/runtime-apis/fetch/) and [execution context](https://developers.cloudflare.com/workers/runtime-apis/context/) |
| Send a later response or new outbound attachment | [Sending Workers API](https://developers.cloudflare.com/email-service/api/send-emails/workers-api/) and [attachment examples](https://developers.cloudflare.com/email-service/examples/email-sending/email-attachments/) |

Plan a single read of raw content when both parsing and storage need it. A queue consumer or later request sends through the outbound API because the original incoming email event is no longer available.

# Email Workers

Use an Email Worker's `email()` handler for custom processing of incoming mail. Use [routing rules](../email-routing/README.md) for simple address-based forwarding. Fetch current documentation before implementing the handler or its dependencies.

| Operation | Documentation |
| --- | --- |
| Forward to a verified destination, reject, or reply within the incoming event | [Email handler API](https://developers.cloudflare.com/email-service/api/route-emails/email-handler/) |
| Send a new message or a later response | [Sending Workers API](https://developers.cloudflare.com/email-service/api/send-emails/workers-api/) |
| Parse and store mail for later processing | [Email storage and processing](https://developers.cloudflare.com/email-service/examples/email-routing/email-storage/) |

`message.raw` is a single-use stream. If parsing and archiving both need the raw content, plan how to reuse it rather than reading the stream twice. Forwarding destinations must be verified; reply requirements are documented separately from outbound sending.

## Reference map

- [Configuration](configuration.md): routing, bindings, local development, and types.
- [API](api.md): message actions, MIME, and sending.
- [Patterns](patterns.md): filtering, storage, attachments, and background processing.
- [Troubleshooting](gotchas.md): stream handling, authentication, limits, and errors.

# Flagship API Reference

## Binding API (Workers)

The binding is available as `env.FLAGS` (type `Flagship` from `@cloudflare/workers-types`).

### Evaluation Methods

All methods are async, never throw, and return the `defaultValue` on errors.

| Method | Signature | Returns |
|--------|-----------|---------|
| `get` | `get(flagKey, defaultValue?, context?)` | `Promise<unknown>` |
| `getBooleanValue` | `getBooleanValue(flagKey, defaultValue, context?)` | `Promise<boolean>` |
| `getStringValue` | `getStringValue(flagKey, defaultValue, context?)` | `Promise<string>` |
| `getNumberValue` | `getNumberValue(flagKey, defaultValue, context?)` | `Promise<number>` |
| `getObjectValue` | `getObjectValue<T>(flagKey, defaultValue, context?)` | `Promise<T>` |
| `getBooleanDetails` | `getBooleanDetails(flagKey, defaultValue, context?)` | `Promise<FlagshipEvaluationDetails<boolean>>` |
| `getStringDetails` | `getStringDetails(flagKey, defaultValue, context?)` | `Promise<FlagshipEvaluationDetails<string>>` |
| `getNumberDetails` | `getNumberDetails(flagKey, defaultValue, context?)` | `Promise<FlagshipEvaluationDetails<number>>` |
| `getObjectDetails` | `getObjectDetails<T>(flagKey, defaultValue, context?)` | `Promise<FlagshipEvaluationDetails<T>>` |

### Parameters (shared across all methods)

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `flagKey` | `string` | Yes | Flag key to evaluate |
| `defaultValue` | varies | Yes (except `get`) | Fallback if evaluation fails or flag not found |
| `context` | `FlagshipEvaluationContext` | No | Attributes for targeting rules (`{ userId: "user-42", country: "US" }`) |

### Types

```typescript
type FlagshipEvaluationContext = Record<string, string | number | boolean>;

interface FlagshipEvaluationDetails<T> {
  flagKey: string;
  value: T;
  variant?: string;     // name of the matched variation
  reason?: string;      // "TARGETING_MATCH" | "DEFAULT" | "DISABLED" | "SPLIT"
  errorCode?: string;   // "TYPE_MISMATCH" | "GENERAL"
  errorMessage?: string;
}
```

### Example

```typescript
export default {
  async fetch(request: Request, env: Env): Promise<Response> {
    const enabled = await env.FLAGS.getBooleanValue("new-feature", false, {
      userId: "user-42",
    });
    return new Response(enabled ? "Feature on" : "Feature off");
  },
};
```

---

## OpenFeature SDK

Package: `@cloudflare/flagship`

### Server Provider (`FlagshipServerProvider`)

For Workers, Node.js, and server-side JavaScript.

**With binding (recommended inside Workers):**

```typescript
import { OpenFeature } from "@openfeature/server-sdk";
import { FlagshipServerProvider } from "@cloudflare/flagship";

await OpenFeature.setProviderAndWait(
  new FlagshipServerProvider({ binding: env.FLAGS }),
);
const client = OpenFeature.getClient();
const enabled = await client.getBooleanValue("new-checkout", false, {
  targetingKey: "user-42",
});
```

**With app ID (Node.js / non-Worker runtimes):**

```typescript
import { OpenFeature } from "@openfeature/server-sdk";
import { FlagshipServerProvider } from "@cloudflare/flagship";

await OpenFeature.setProviderAndWait(
  new FlagshipServerProvider({
    appId: "<APP_ID>",
    accountId: "<ACCOUNT_ID>",
    authToken: "<API_TOKEN>",
  }),
);
const client = OpenFeature.getClient();
const enabled = await client.getBooleanValue("new-checkout", false, {
  targetingKey: "user-42",
});
```

### Client Provider (`FlagshipClientProvider`)

For browser applications. Pre-fetches flags on init, evaluates synchronously.

```typescript
import { OpenFeature } from "@openfeature/web-sdk";
import { FlagshipClientProvider } from "@cloudflare/flagship";

await OpenFeature.setProviderAndWait(
  new FlagshipClientProvider({
    appId: "<APP_ID>",
    accountId: "<ACCOUNT_ID>",
    authToken: "<API_TOKEN>",
    prefetchFlags: ["promo-banner", "dark-mode"],
  }),
);
await OpenFeature.setContext({ targetingKey: "user-42", plan: "enterprise" });
const client = OpenFeature.getClient();

// Synchronous — no await needed
const showBanner = client.getBooleanValue("promo-banner", false);
```

**Important:** Only flags listed in `prefetchFlags` are available. Unlisted flags return `FLAG_NOT_FOUND`.

### SDK Hooks

```typescript
import { LoggingHook, TelemetryHook } from "@cloudflare/flagship";
OpenFeature.addHooks(new LoggingHook(), new TelemetryHook());
```

---

## REST API (Flag Management)

Source of truth: [Cloudflare Flagship API reference](https://developers.cloudflare.com/api/resources/flagship/). Use it to verify REST paths, envelopes, response fields, and permission wording before relying on examples here.

### FIRST: Check Prerequisites

Before making any REST API calls (create, read, update, delete, toggle flags), verify these environment variables are set:

| Variable | Purpose | How to get |
|----------|---------|------------|
| `CLOUDFLARE_ACCOUNT_ID` | Account identifier | Dashboard URL or `wrangler whoami` |
| `CLOUDFLARE_API_TOKEN` | Bearer token for API auth | [Create API token](https://dash.cloudflare.com/profile/api-tokens) with Flagship permissions |
| `FLAGSHIP_APP_ID` | Target app UUID | Dashboard under **Compute > Flagship**, or `GET /apps` endpoint |

Check with:

```bash
echo "CLOUDFLARE_ACCOUNT_ID=${CLOUDFLARE_ACCOUNT_ID:-(not set)}"
echo "CLOUDFLARE_API_TOKEN=${CLOUDFLARE_API_TOKEN:-(not set)}"
echo "FLAGSHIP_APP_ID=${FLAGSHIP_APP_ID:-(not set)}"
```

**If any are missing, ask the user to provide them before proceeding.**

### Base URL and Auth

Base URL: `https://api.cloudflare.com/client/v4/accounts/{account_id}/flagship`

Authentication: `Authorization: Bearer <API_TOKEN>`

Management endpoints use the Cloudflare v4 envelope. On success, the payload is under `result`; errors are an array under `errors`.

```jsonc
// Success
{ "success": true, "result": <T>, "errors": [], "messages": [] }

// Paginated success
{
  "success": true,
  "result": [<T>],
  "result_info": { "count": 50, "cursor": "next-cursor-or-null" },
  "errors": [],
  "messages": []
}

// Error
{ "success": false, "result": null, "errors": [{ "message": "message" }], "messages": [] }
```

### App Endpoints

| Method | Path | Description |
|--------|------|-------------|
| `GET` | `/apps` | List all apps |
| `GET` | `/apps/{app_id}` | Get app |
| `POST` | `/apps` | Create app (`{ "name": "my-app" }`) |
| `PUT` | `/apps/{app_id}` | Update app (`{ "name": "new-name" }`) |
| `DELETE` | `/apps/{app_id}` | Delete app |

App name constraints: alphanumeric + hyphens + underscores, 1-64 chars.

### Flag Endpoints

| Method | Path | Description |
|--------|------|-------------|
| `GET` | `/apps/{app_id}/flags?limit=50&cursor=<cursor>` | List flags (paginated) |
| `GET` | `/apps/{app_id}/flags/{flag_key}` | Get flag |
| `POST` | `/apps/{app_id}/flags` | Create flag |
| `PUT` | `/apps/{app_id}/flags/{flag_key}` | Update flag (full replace) |
| `DELETE` | `/apps/{app_id}/flags/{flag_key}` | Delete flag |
| `GET` | `/apps/{app_id}/flags/{flag_key}/changelog?limit=20&cursor=<cursor>` | Flag changelog |

### Evaluate Endpoint

```
GET /apps/{app_id}/evaluate?flagKey=<key>&<context-attrs>
```

Requires an API token with the `com.cloudflare.account.flagship.evaluate` permission. Context attributes passed as query params. This endpoint is not wrapped in the management envelope; the SDK contract returns OpenFeature-style camelCase:

```json
{
  "flagKey": "my-flag",
  "value": true,
  "variant": "on",
  "reason": "SPLIT"
}
```

Reasons: `TARGETING_MATCH`, `SPLIT`, `DEFAULT`, `DISABLED`.

### Management Response Payloads

Management endpoints are wrapped in the Cloudflare v4 envelope shown above. Common `.result` payloads:

**App result**

```json
{
  "id": "app-uuid",
  "name": "my-app",
  "created_at": "2026-06-09T12:00:00.000Z",
  "updated_at": "2026-06-09T12:00:00.000Z",
  "updated_by": "user@example.com"
}
```

**Flag result**

```json
{
  "key": "my-flag",
  "type": "boolean",
  "default_variation": "off",
  "variations": { "on": true, "off": false },
  "rules": [],
  "description": "Enables the new feature",
  "enabled": true,
  "updated_at": "2026-06-09T12:00:00.000Z",
  "updated_by": "user@example.com"
}
```

**Changelog entry**

```json
{
  "flag_key": "my-flag",
  "event": "update",
  "after": { "key": "my-flag", "default_variation": "off", "variations": { "on": true, "off": false }, "rules": [], "enabled": true },
  "diff": { "enabled": { "from": false, "to": true } }
}
```

Changelog entries include the full flag state after the change. `update` entries also include `diff`.

---

## FlagDefinition Schema

```json
{
  "key": "my-flag",
  "type": "boolean",
  "default_variation": "off",
  "variations": {
    "on": true,
    "off": false
  },
  "rules": [
    {
      "priority": 1,
      "conditions": [
        {
          "attribute": "email",
          "operator": "ends_with",
          "value": "@cloudflare.com"
        }
      ],
      "serve_variation": "on",
      "rollout": { "percentage": 100 }
    }
  ],
  "description": "Enables the new feature",
  "enabled": true
}
```

### Field Constraints

| Field | Type | Constraints |
|-------|------|-------------|
| `key` | string | 1-64 chars, `/^[a-zA-Z0-9_-]+$/` |
| `type` | enum | Optional. `boolean`, `string`, `number`, `json` (auto-inferred from variations) |
| `default_variation` | string | Must be a key in `variations` |
| `variations` | `Record<string, T>` | At least one. All values same type. Keys: alphanumeric/hyphens/underscores, max 64 chars. Values max 10KB. |
| `rules` | `Rule[]` | Can be empty. No duplicate priorities. |
| `description` | string? | Max 512 chars, nullable |
| `enabled` | boolean | Required. `false` = always returns default variation. |

### Rule Schema

```json
{
  "priority": 1,
  "conditions": [ /* Condition[] */ ],
  "serve_variation": "on",
  "rollout": { "percentage": 50, "attribute": "targetingKey" }
}
```

- `priority`: integer >= 1, unique across rules in the flag (lower = evaluated first)
- `conditions`: array of base or logical conditions
- `serve_variation`: must be a key in `variations`
- `rollout`: optional. `percentage` 0-100. `attribute` defaults to `targetingKey`.

### Condition Schema

**Base condition:**

```json
{ "attribute": "email", "operator": "ends_with", "value": "@cloudflare.com" }
```

**Logical condition (AND/OR):**

```json
{
  "logical_operator": "AND",
  "clauses": [
    { "attribute": "country", "operator": "equals", "value": "US" },
    { "attribute": "plan", "operator": "in", "value": ["enterprise", "business"] }
  ]
}
```

Nesting supported up to 6 levels deep.

### Operators

| Operator | Description | Value Type |
|----------|-------------|------------|
| `equals` | Exact match (case-sensitive) | String |
| `not_equals` | Not exact match | String |
| `greater_than` | Numeric / datetime > | Number, ISO 8601 |
| `less_than` | Numeric / datetime < | Number, ISO 8601 |
| `greater_than_or_equals` | >= | Number, ISO 8601 |
| `less_than_or_equals` | <= | Number, ISO 8601 |
| `contains` | Substring match (case-sensitive) | String |
| `starts_with` | Prefix match | String |
| `ends_with` | Suffix match | String |
| `in` | Value in array | Array |
| `not_in` | Value not in array | Array |

---

## Rate Limits

| Operation | Limit |
|-----------|-------|
| Mutations (POST/PUT/DELETE) | 60 per 60s per account:app |
| Reads (GET) | 600 per 60s per account:app |

## Error Codes

| HTTP Status | Meaning |
|-------------|---------|
| 200 | Success (read/update/delete) |
| 201 | Created (create) |
| 400 | Validation error (check `errors[].message`) |
| 401 | Invalid or missing token |
| 404 | Flag or app not found |
| 409 | Flag key already exists (create) |
| 429 | Rate limited |

# Flagship Configuration

## Wrangler Binding Setup

Add a Flagship binding to your Wrangler config to access flags via `env.FLAGS`.

### Single App

```jsonc
// wrangler.jsonc
{
  "flagship": {
    "binding": "FLAGS",
    "app_id": "<APP_ID>"
  }
}
```

```toml
# wrangler.toml
[flagship]
binding = "FLAGS"
app_id = "<APP_ID>"
```

### Multiple Apps

```jsonc
// wrangler.jsonc
{
  "flagship": [
    {
      "binding": "FLAGS",
      "app_id": "<APP_ID_1>"
    },
    {
      "binding": "EXPERIMENT_FLAGS",
      "app_id": "<APP_ID_2>"
    }
  ]
}
```

```toml
# wrangler.toml
[[flagship]]
binding = "FLAGS"
app_id = "<APP_ID_1>"

[[flagship]]
binding = "EXPERIMENT_FLAGS"
app_id = "<APP_ID_2>"
```

### Generate Types

After adding the binding, generate TypeScript types:

```bash
npx wrangler types
```

This creates the `Env` interface with each binding typed as `Flagship`:

```typescript
interface Env {
  FLAGS: Flagship;
  EXPERIMENT_FLAGS: Flagship; // if multiple
}
```

The `Flagship` type comes from `@cloudflare/workers-types`.

---

## OpenFeature SDK Installation

### Server-Side (Workers, Node.js)

```bash
npm i @cloudflare/flagship @openfeature/server-sdk
```

### Browser

```bash
npm i @cloudflare/flagship @openfeature/web-sdk
```

---

## SDK Provider Setup

### Server Provider — With Binding (Workers)

Recommended approach inside Workers. No HTTP overhead, auth handled automatically.

```typescript
import { OpenFeature } from "@openfeature/server-sdk";
import { FlagshipServerProvider } from "@cloudflare/flagship";

export default {
  async fetch(request: Request, env: Env): Promise<Response> {
    await OpenFeature.setProviderAndWait(
      new FlagshipServerProvider({ binding: env.FLAGS }),
    );
    const client = OpenFeature.getClient();
    // ... evaluate flags
  },
};
```

### Server Provider — With App ID (Node.js)

For non-Worker runtimes. Requires an API token with Flagship read permissions.

```typescript
import { OpenFeature } from "@openfeature/server-sdk";
import { FlagshipServerProvider } from "@cloudflare/flagship";

await OpenFeature.setProviderAndWait(
  new FlagshipServerProvider({
    appId: "<APP_ID>",
    accountId: "<ACCOUNT_ID>",
    authToken: "<API_TOKEN>",
  }),
);
const client = OpenFeature.getClient();
```

### Client Provider (Browser)

Pre-fetches flags on init, then evaluates synchronously. Only `prefetchFlags` are available.

```typescript
import { OpenFeature } from "@openfeature/web-sdk";
import { FlagshipClientProvider } from "@cloudflare/flagship";

await OpenFeature.setProviderAndWait(
  new FlagshipClientProvider({
    appId: "<APP_ID>",
    accountId: "<ACCOUNT_ID>",
    authToken: "<API_TOKEN>",
    prefetchFlags: ["promo-banner", "dark-mode", "max-uploads"],
  }),
);
await OpenFeature.setContext({ targetingKey: "user-42", plan: "enterprise" });
const client = OpenFeature.getClient();
```

### Provider Options Reference

**FlagshipServerProvider:**

| Option | Type | Required | Description |
|--------|------|----------|-------------|
| `binding` | `Flagship` | No | Binding from `env.FLAGS`. Use inside Workers. |
| `appId` | string | No | App ID from dashboard. Required without binding. |
| `accountId` | string | No | Cloudflare account ID. Required without binding. |
| `authToken` | string | No | API token with Flagship read permissions. Required without binding. |

Provide either `binding` or all three of `appId` + `accountId` + `authToken`.

**FlagshipClientProvider:**

| Option | Type | Required | Description |
|--------|------|----------|-------------|
| `appId` | string | Yes | App ID from dashboard |
| `accountId` | string | Yes | Cloudflare account ID |
| `authToken` | string | Yes | API token with Flagship read permissions |
| `prefetchFlags` | string[] | Yes | Flag keys to prefetch. Unlisted flags return `FLAG_NOT_FOUND`. |

---

## REST API Authentication

For managing flags via the REST API (create, update, delete), set these environment variables:

| Variable | Description |
|----------|-------------|
| `CLOUDFLARE_ACCOUNT_ID` | Your Cloudflare account ID |
| `CLOUDFLARE_API_TOKEN` | API token with Flagship permissions |
| `FLAGSHIP_APP_ID` | Target app UUID (from dashboard under **Compute > Flagship**, or `GET /apps`) |

Base URL: `https://api.cloudflare.com/client/v4/accounts/$CLOUDFLARE_ACCOUNT_ID/flagship`

```bash
curl -s -H "Authorization: Bearer $CLOUDFLARE_API_TOKEN" \
  "https://api.cloudflare.com/client/v4/accounts/$CLOUDFLARE_ACCOUNT_ID/flagship/apps" | jq .
```

App IDs are shown in the Cloudflare dashboard under **Compute > Flagship**.

---

## Local Development

Flagship bindings work in local dev with `wrangler dev`. Flag evaluation uses the live Flagship configuration — there is no local flag store. Ensure the `app_id` in your Wrangler config points to a valid app.

```bash
npx wrangler dev
```

# Flagship Gotchas & Troubleshooting

## Common Errors

### Flag Always Returns Default Value

**Cause:** Flag is disabled (`enabled: false`), or no targeting rules match, or evaluation context is missing expected attributes.

**Solution:** Check these in order:

1. Is the flag enabled? (`"enabled": true`)
2. Do your targeting rules match the context you're passing?
3. Are you passing the right attributes in the evaluation context?

```typescript
// ❌ BAD — no context, rules can't match
const val = await env.FLAGS.getBooleanValue("my-flag", false);

// ✅ GOOD — pass context attributes that rules reference
const val = await env.FLAGS.getBooleanValue("my-flag", false, {
  userId: "user-42",
  plan: "enterprise",
});
```

### TYPE_MISMATCH Error in Details

**Cause:** Calling a typed method on a flag with a different type (e.g., `getBooleanValue` on a string flag).

**Solution:** Use the method matching the flag's variation type.

```typescript
// ❌ BAD — flag "checkout-flow" has string variations
const val = await env.FLAGS.getBooleanValue("checkout-flow", false);

// ✅ GOOD
const val = await env.FLAGS.getStringValue("checkout-flow", "original");
```

### 409 Conflict on Flag Creation

**Cause:** A flag with that key already exists in the app.

**Solution:** Use a different key, or GET + PUT to update the existing flag.

### Inconsistent Rollout Results

**Cause:** `targetingKey` (or the configured bucketing attribute) is missing from the evaluation context, causing random bucketing on each request.

**Solution:** Always pass a stable identifier:

```typescript
// ❌ BAD — no targetingKey, rollout is random per request
const val = await env.FLAGS.getBooleanValue("gradual-rollout", false);

// ✅ GOOD — stable userId for consistent bucketing
const val = await env.FLAGS.getBooleanValue("gradual-rollout", false, {
  userId: sessionUserId,
});
```

### Update Overwrites Entire Flag

**Cause:** PUT replaces the full `FlagDefinition`. Sending only changed fields deletes the rest.

**Solution:** Always read-modify-write:

```bash
# ❌ BAD — overwrites the entire flag, losing rules/variations
curl -X PUT -d '{"enabled": true}' ...

# ✅ GOOD — GET first, modify, PUT back
FLAG=$(curl -s -H "Authorization: Bearer $TOKEN" "$URL/flags/my-flag" | jq '.result')
UPDATED=$(echo "$FLAG" | jq '.enabled = true')
echo "$UPDATED" | curl -s -X PUT -H "Authorization: Bearer $TOKEN" -H "Content-Type: application/json" -d @- "$URL/flags/my-flag"
```

### Reading REST Envelope Fields

**Cause:** Management endpoints use Cloudflare v4 envelopes, not raw payloads.

**Solution:** Read `.result` for successful payloads, `.result_info.cursor` for pagination, and `.errors[].message` for errors.

```bash
jq '.result'
jq '.result_info.cursor'
jq '.errors[].message'
```

### Mixing CamelCase and Snake Case in REST Responses

**Cause:** Management API responses are public API JSON and use snake_case. Evaluation responses use OpenFeature-style camelCase.

**Solution:** For management endpoints use `default_variation`, `serve_variation`, `updated_at`, `updated_by`, and changelog `flag_key`. For `/evaluate`, use `flagKey`, `variant`, and `reason`.

### FLAG_NOT_FOUND in Client Provider

**Cause:** Flag key not included in `prefetchFlags` array.

**Solution:** Add the flag key to `prefetchFlags` when initializing `FlagshipClientProvider`.

### Client Provider Token Exposure

**Cause:** The `authToken` passed to `FlagshipClientProvider` is visible in the browser. It can evaluate flags across all apps in the account.

**Solution:** Use a token with minimal permissions (Flagship Evaluate only). Never use a token with write/management permissions in the browser.

---

## Limits

| Limit | Value | Notes |
|-------|-------|-------|
| Flag key length | 1-64 chars | Alphanumeric, hyphens, underscores only |
| Flag key pattern | `/^[a-zA-Z0-9_-]+$/` | — |
| Variation value size | 10KB max | Per variation, serialized |
| Variation name length | 64 chars max | Alphanumeric, hyphens, underscores |
| Description length | 512 chars max | Nullable |
| App name length | 1-64 chars | Alphanumeric, hyphens, underscores |
| Logical nesting depth | 6 levels | AND/OR conditions |
| Mutation rate limit | 60 / 60s | Per account:app |
| Read rate limit | 600 / 60s | Per account:app |
| Rollout percentage | 0-100 | Integer |
| Rule priorities | Unique integers >= 1 | Lower = evaluated first |

---

## Anti-Patterns

### Evaluating Flags in a Tight Loop

Flag evaluation via the binding is fast but not free. Avoid evaluating the same flag repeatedly in a loop — evaluate once and reuse the result.

```typescript
// ❌ BAD
for (const item of items) {
  const enabled = await env.FLAGS.getBooleanValue("my-flag", false, ctx);
  // ...
}

// ✅ GOOD
const enabled = await env.FLAGS.getBooleanValue("my-flag", false, ctx);
for (const item of items) {
  // use `enabled`
}
```

### Using the SDK Inside Workers When Binding Is Available

The binding avoids HTTP overhead entirely. Only use the SDK inside Workers when you specifically need OpenFeature vendor-neutrality.

```typescript
// ❌ Unnecessary HTTP overhead inside a Worker
const provider = new FlagshipServerProvider({
  appId: "...", accountId: "...", authToken: "...",
});

// ✅ Use the binding directly, or pass it to the SDK
const provider = new FlagshipServerProvider({ binding: env.FLAGS });
```

### Partial PUT Updates

The flag update API (PUT) requires the complete `FlagDefinition`. Sending only changed fields silently drops everything else. Always GET first, then modify and PUT back the full object.

### Stale Flag Cleanup

Flags that are disabled and no longer referenced in code should be deleted. Stale flags clutter the dashboard and make it harder to understand which flags are active. Follow the safe deletion workflow in `patterns.md`.

---

## Propagation Behavior

Flag changes propagate globally within seconds. During the brief propagation window, some regions may serve the previous value. After propagation completes, all evaluations return the updated value.

- No Worker redeployment needed for flag changes.
- If the dashboard is temporarily unavailable, evaluation continues using the last propagated configuration.
- Flag changes made via the REST API and dashboard are equivalent — both trigger propagation.

# Flagship Patterns & Best Practices

## Evaluating Flags in Workers (Binding)

### Simple Boolean Toggle

```typescript
export default {
  async fetch(request: Request, env: Env): Promise<Response> {
    const showNewUI = await env.FLAGS.getBooleanValue("new-ui", false, {
      userId: "user-42",
    });

    if (showNewUI) {
      return new Response("New UI");
    }
    return new Response("Classic UI");
  },
};
```

### Multi-Variant String Flag

```typescript
const checkoutFlow = await env.FLAGS.getStringValue(
  "checkout-flow",
  "original",
  { userId, country: "US" },
);

switch (checkoutFlow) {
  case "streamlined":
    return handleStreamlined(request);
  case "one-click":
    return handleOneClick(request);
  default:
    return handleOriginal(request);
}
```

### JSON Config Flag

```typescript
interface RateLimitConfig {
  rpm: number;
  burst: number;
}

const limits = await env.FLAGS.getObjectValue<RateLimitConfig>(
  "rate-limits",
  { rpm: 100, burst: 20 },
  { plan: userPlan },
);
```

### Using Details for Observability

```typescript
const details = await env.FLAGS.getBooleanDetails("new-checkout", false, {
  userId: "user-42",
});

console.log(details.value);     // true
console.log(details.variant);   // "on"
console.log(details.reason);    // "TARGETING_MATCH"
console.log(details.errorCode); // undefined (no error)
```

---

## Evaluating Flags with OpenFeature (Workers)

### Binding Passthrough (Recommended)

```typescript
import { OpenFeature } from "@openfeature/server-sdk";
import { FlagshipServerProvider } from "@cloudflare/flagship";

export default {
  async fetch(request: Request, env: Env): Promise<Response> {
    await OpenFeature.setProviderAndWait(
      new FlagshipServerProvider({ binding: env.FLAGS }),
    );
    const client = OpenFeature.getClient();

    const enabled = await client.getBooleanValue("new-checkout", false, {
      targetingKey: "user-42",
      plan: "enterprise",
      country: "US",
    });

    return new Response(enabled ? "New checkout" : "Standard checkout");
  },
};
```

### Migration from Another Provider

Only the provider initialization changes — evaluation call sites stay the same:

```typescript
// ❌ Before (LaunchDarkly)
await OpenFeature.setProviderAndWait(
  new LaunchDarklyProvider({ sdkKey: "..." }),
);

// ✅ After (Flagship)
await OpenFeature.setProviderAndWait(
  new FlagshipServerProvider({ binding: env.FLAGS }),
);

// Evaluation code is unchanged
const enabled = await client.getBooleanValue("my-flag", false, {
  targetingKey: "user-42",
});
```

---

## Managing Flags via REST API

All examples use `api.cloudflare.com`. Set `CLOUDFLARE_ACCOUNT_ID`, `FLAGSHIP_APP_ID`, and `CLOUDFLARE_API_TOKEN` first.

### Create a Boolean Flag

```bash
curl -s -X POST \
  -H "Authorization: Bearer $CLOUDFLARE_API_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "key": "new-feature",
    "default_variation": "off",
    "variations": { "on": true, "off": false },
    "rules": [],
    "description": "Enable the new feature",
    "enabled": false
  }' \
  "https://api.cloudflare.com/client/v4/accounts/$CLOUDFLARE_ACCOUNT_ID/flagship/apps/$FLAGSHIP_APP_ID/flags" | jq .
```

### Create a Flag with Internal-Only Targeting

```bash
curl -s -X POST \
  -H "Authorization: Bearer $CLOUDFLARE_API_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "key": "beta-feature",
    "default_variation": "off",
    "variations": { "on": true, "off": false },
    "rules": [
      {
        "priority": 1,
        "conditions": [
          { "attribute": "email", "operator": "ends_with", "value": "@cloudflare.com" }
        ],
        "serve_variation": "on"
      }
    ],
    "description": "Beta feature for internal users",
    "enabled": true
  }' \
  "https://api.cloudflare.com/client/v4/accounts/$CLOUDFLARE_ACCOUNT_ID/flagship/apps/$FLAGSHIP_APP_ID/flags" | jq .
```

### Create a JSON Config Flag

```bash
curl -s -X POST \
  -H "Authorization: Bearer $CLOUDFLARE_API_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "key": "rate-limits",
    "default_variation": "standard",
    "variations": {
      "standard": { "rpm": 100, "burst": 20 },
      "premium": { "rpm": 1000, "burst": 200 }
    },
    "rules": [
      {
        "priority": 1,
        "conditions": [
          { "attribute": "plan", "operator": "in", "value": ["enterprise", "business"] }
        ],
        "serve_variation": "premium"
      }
    ],
    "description": "Rate limit configuration by plan",
    "enabled": true
  }' \
  "https://api.cloudflare.com/client/v4/accounts/$CLOUDFLARE_ACCOUNT_ID/flagship/apps/$FLAGSHIP_APP_ID/flags" | jq .
```

### Read a Flag

```bash
curl -s -H "Authorization: Bearer $CLOUDFLARE_API_TOKEN" \
  "https://api.cloudflare.com/client/v4/accounts/$CLOUDFLARE_ACCOUNT_ID/flagship/apps/$FLAGSHIP_APP_ID/flags/new-feature" | jq .
```

### List All Flags (with pagination)

```bash
curl -s -H "Authorization: Bearer $CLOUDFLARE_API_TOKEN" \
  "https://api.cloudflare.com/client/v4/accounts/$CLOUDFLARE_ACCOUNT_ID/flagship/apps/$FLAGSHIP_APP_ID/flags?limit=50" | jq .
```

If `result_info.cursor` is non-null, fetch the next page:

```bash
curl -s -H "Authorization: Bearer $CLOUDFLARE_API_TOKEN" \
  "https://api.cloudflare.com/client/v4/accounts/$CLOUDFLARE_ACCOUNT_ID/flagship/apps/$FLAGSHIP_APP_ID/flags?limit=50&cursor=<cursor>" | jq .
```

### Update a Flag (Full Replace)

Updates use PUT with the full `FlagDefinition`. Always GET first, modify, then PUT back.

```bash
# 1. Read current flag
FLAG=$(curl -s -H "Authorization: Bearer $CLOUDFLARE_API_TOKEN" \
  "https://api.cloudflare.com/client/v4/accounts/$CLOUDFLARE_ACCOUNT_ID/flagship/apps/$FLAGSHIP_APP_ID/flags/new-feature" | jq '.result')

# 2. Modify (e.g., enable the flag)
UPDATED=$(echo "$FLAG" | jq '.enabled = true')

# 3. PUT back
echo "$UPDATED" | curl -s -X PUT \
  -H "Authorization: Bearer $CLOUDFLARE_API_TOKEN" \
  -H "Content-Type: application/json" \
  -d @- \
  "https://api.cloudflare.com/client/v4/accounts/$CLOUDFLARE_ACCOUNT_ID/flagship/apps/$FLAGSHIP_APP_ID/flags/new-feature" | jq .
```

### Toggle a Flag On

Read-modify-write to set `enabled: true`:

```bash
BASE="https://api.cloudflare.com/client/v4/accounts/$CLOUDFLARE_ACCOUNT_ID/flagship/apps/$FLAGSHIP_APP_ID/flags"

FLAG=$(curl -s -H "Authorization: Bearer $CLOUDFLARE_API_TOKEN" "$BASE/new-feature" | jq '.result')
UPDATED=$(echo "$FLAG" | jq '.enabled = true')
echo "$UPDATED" | curl -s -X PUT \
  -H "Authorization: Bearer $CLOUDFLARE_API_TOKEN" \
  -H "Content-Type: application/json" \
  -d @- "$BASE/new-feature" | jq .
```

### Toggle a Flag Off (Disable)

Same pattern, set `enabled: false`. The flag immediately returns its default variation for all evaluations.

```bash
BASE="https://api.cloudflare.com/client/v4/accounts/$CLOUDFLARE_ACCOUNT_ID/flagship/apps/$FLAGSHIP_APP_ID/flags"

FLAG=$(curl -s -H "Authorization: Bearer $CLOUDFLARE_API_TOKEN" "$BASE/new-feature" | jq '.result')
UPDATED=$(echo "$FLAG" | jq '.enabled = false')
echo "$UPDATED" | curl -s -X PUT \
  -H "Authorization: Bearer $CLOUDFLARE_API_TOKEN" \
  -H "Content-Type: application/json" \
  -d @- "$BASE/new-feature" | jq .
```

### Add a Targeting Rule to an Existing Flag

Append a rule to the existing rules array. Pick a priority that doesn't collide with existing rules.

```bash
BASE="https://api.cloudflare.com/client/v4/accounts/$CLOUDFLARE_ACCOUNT_ID/flagship/apps/$FLAGSHIP_APP_ID/flags"

FLAG=$(curl -s -H "Authorization: Bearer $CLOUDFLARE_API_TOKEN" "$BASE/new-feature" | jq '.result')
UPDATED=$(echo "$FLAG" | jq '.rules += [{
  "priority": 2,
  "conditions": [{ "attribute": "plan", "operator": "equals", "value": "enterprise" }],
  "serve_variation": "on"
}]')
echo "$UPDATED" | curl -s -X PUT \
  -H "Authorization: Bearer $CLOUDFLARE_API_TOKEN" \
  -H "Content-Type: application/json" \
  -d @- "$BASE/new-feature" | jq .
```

### Change Rollout Percentage

Update the rollout percentage on an existing rule (e.g., rule at index 0):

```bash
BASE="https://api.cloudflare.com/client/v4/accounts/$CLOUDFLARE_ACCOUNT_ID/flagship/apps/$FLAGSHIP_APP_ID/flags"

FLAG=$(curl -s -H "Authorization: Bearer $CLOUDFLARE_API_TOKEN" "$BASE/gradual-rollout" | jq '.result')
UPDATED=$(echo "$FLAG" | jq '.rules[0].rollout.percentage = 50')
echo "$UPDATED" | curl -s -X PUT \
  -H "Authorization: Bearer $CLOUDFLARE_API_TOKEN" \
  -H "Content-Type: application/json" \
  -d @- "$BASE/gradual-rollout" | jq .
```

### Change Default Variation

```bash
BASE="https://api.cloudflare.com/client/v4/accounts/$CLOUDFLARE_ACCOUNT_ID/flagship/apps/$FLAGSHIP_APP_ID/flags"

FLAG=$(curl -s -H "Authorization: Bearer $CLOUDFLARE_API_TOKEN" "$BASE/new-feature" | jq '.result')
UPDATED=$(echo "$FLAG" | jq '.default_variation = "on"')
echo "$UPDATED" | curl -s -X PUT \
  -H "Authorization: Bearer $CLOUDFLARE_API_TOKEN" \
  -H "Content-Type: application/json" \
  -d @- "$BASE/new-feature" | jq .
```

### Add a New Variation

```bash
BASE="https://api.cloudflare.com/client/v4/accounts/$CLOUDFLARE_ACCOUNT_ID/flagship/apps/$FLAGSHIP_APP_ID/flags"

FLAG=$(curl -s -H "Authorization: Bearer $CLOUDFLARE_API_TOKEN" "$BASE/checkout-flow" | jq '.result')
UPDATED=$(echo "$FLAG" | jq '.variations["treatment-c"] = "minimal"')
echo "$UPDATED" | curl -s -X PUT \
  -H "Authorization: Bearer $CLOUDFLARE_API_TOKEN" \
  -H "Content-Type: application/json" \
  -d @- "$BASE/checkout-flow" | jq .
```

### Remove a Rule

Remove a rule by filtering on priority:

```bash
BASE="https://api.cloudflare.com/client/v4/accounts/$CLOUDFLARE_ACCOUNT_ID/flagship/apps/$FLAGSHIP_APP_ID/flags"

FLAG=$(curl -s -H "Authorization: Bearer $CLOUDFLARE_API_TOKEN" "$BASE/new-feature" | jq '.result')
UPDATED=$(echo "$FLAG" | jq '.rules = [.rules[] | select(.priority != 2)]')
echo "$UPDATED" | curl -s -X PUT \
  -H "Authorization: Bearer $CLOUDFLARE_API_TOKEN" \
  -H "Content-Type: application/json" \
  -d @- "$BASE/new-feature" | jq .
```

### Delete a Flag

```bash
curl -s -X DELETE \
  -H "Authorization: Bearer $CLOUDFLARE_API_TOKEN" \
  "https://api.cloudflare.com/client/v4/accounts/$CLOUDFLARE_ACCOUNT_ID/flagship/apps/$FLAGSHIP_APP_ID/flags/old-feature" | jq .
```

---

## Targeting Rule Patterns

### Enterprise-Only Access

```json
{
  "priority": 1,
  "conditions": [
    { "attribute": "plan", "operator": "equals", "value": "enterprise" }
  ],
  "serve_variation": "on"
}
```

### Country-Based Targeting with Logical AND/OR

Target enterprise users in the US or Canada:

```json
{
  "priority": 1,
  "conditions": [
    {
      "logical_operator": "AND",
      "clauses": [
        { "attribute": "plan", "operator": "equals", "value": "enterprise" },
        {
          "logical_operator": "OR",
          "clauses": [
            { "attribute": "country", "operator": "equals", "value": "US" },
            { "attribute": "country", "operator": "equals", "value": "CA" }
          ]
        }
      ]
    }
  ],
  "serve_variation": "on"
}
```

### Percentage Rollout

Gradually roll out to 10% of users:

```json
{
  "priority": 1,
  "conditions": [
    { "attribute": "targetingKey", "operator": "not_equals", "value": "" }
  ],
  "serve_variation": "on",
  "rollout": {
    "percentage": 10,
    "attribute": "targetingKey"
  }
}
```

### A/B/n (Multi-Variant) Testing

To split traffic across N variants, create one rule per variant with **cumulative** rollout percentages. Flagship evaluates rules in priority order. If a rule's conditions match but the user misses that rule's rollout percentage, evaluation continues to the next rule. Use the same stable rollout attribute on every rule so each user is compared against the same bucket as the thresholds increase.

The example uses `conditions: []` because the rules are intended to match every context. For sticky user assignment, callers must still pass the configured bucketing attribute (`targetingKey` here); otherwise Flagship uses a random bucket per request.

For example, to split traffic 30% / 40% / 30% across variants A, B, and C:

| Variant | Share | Cumulative threshold |
|---------|-------|----------------------|
| A       | 30%   | 30                   |
| B       | 40%   | 70                   |
| C       | 30%   | 100                  |

```json
"rules": [
  {
    "priority": 1,
    "conditions": [],
    "serve_variation": "variant-a",
    "rollout": { "percentage": 30, "attribute": "targetingKey" }
  },
  {
    "priority": 2,
    "conditions": [],
    "serve_variation": "variant-b",
    "rollout": { "percentage": 70, "attribute": "targetingKey" }
  },
  {
    "priority": 3,
    "conditions": [],
    "serve_variation": "variant-c",
    "rollout": { "percentage": 100, "attribute": "targetingKey" }
  }
]
```

Key points:
- Rules are evaluated lowest-priority-number first. A user who falls into rule 1's 0-30% bucket gets `variant-a` and is not evaluated further.
- Rule 2's 70% threshold covers the next 40% of users (31-70%).
- Rule 3's 100% threshold catches the remaining 30% (71-100%).
- Always set the last rule to `100` so every context with the bucketing attribute is assigned a variant.
- For sticky A/B/n assignment, pass a stable `targetingKey` or configured bucketing attribute. Without it, rollout assignment is random per request, which can be useful for request-level sampling but is usually wrong for user experiments.
- A percentage rollout match reports reason `SPLIT` in evaluation details.

### Progressive Rollout Workflow

1. Create flag with 5% rollout, enable it
2. Monitor metrics
3. Increase to 25% → 50% → 100% by updating the `rollout.percentage`
4. Once at 100%, remove the rule and set `default_variation` to the winning variation
5. Eventually remove the flag and the code branch

---

## Safe Deletion Workflow

1. **Disable** the flag first (`enabled: false`) — confirms nothing depends on it being active
2. **Monitor** for unexpected behavior
3. **Remove** flag evaluation code from your application
4. **Deploy** the code change
5. **Delete** the flag via API

# Cloudflare Flagship

Feature flag service for controlling feature visibility without redeploying code. Define flags with targeting rules and percentage-based rollouts, then evaluate them in Workers via a native binding or from any JavaScript runtime via the OpenFeature SDK.

## When to Use

| Need | Use Flagship? | Alternative |
|------|--------------|-------------|
| Feature toggles (on/off) | Yes | — |
| Gradual rollouts (percentage-based) | Yes | — |
| A/B testing with attribute targeting | Yes | — |
| Multi-variant configuration delivery | Yes | — |
| Environment-specific config (dev/staging/prod) | Consider | Wrangler environments, secrets |
| Static config that never changes | No | `wrangler.jsonc` vars |
| Per-request rate limiting | No | Rate Limiting rules |

## Key Concepts

- **Apps** — Top-level organizational unit. Maps to a project or service. Each account can have multiple apps.
- **Flags** — Named feature toggles with a key, variations, targeting rules, and enabled/disabled state.
- **Variations** — Possible values a flag returns. Types: boolean, string, number, JSON object. All variations on a flag must share the same type.
- **Targeting rules** — Sequential, priority-ordered conditions that determine which variation to serve. First match wins; no match returns the default.
- **Evaluation context** — Key-value attributes (`userId`, `country`, `plan`, etc.) passed at evaluation time for rule matching and rollout bucketing.
- **Percentage rollouts** — Gradually release to a fraction of users. Consistent hashing on a configurable attribute ensures sticky bucketing.

## Two Evaluation Paths

| Path | Runtime | Package | Latency | Auth |
|------|---------|---------|---------|------|
| **Binding** (`env.FLAGS`) | Workers only | `@cloudflare/workers-types` | Lowest (no HTTP) | Automatic via binding |
| **OpenFeature SDK** | Workers, Node.js, browser | `@cloudflare/flagship` + `@openfeature/server-sdk` or `@openfeature/web-sdk` | HTTP per eval (server) or prefetch (client) | API token or binding passthrough |

**Recommendation:** Use the binding inside Workers. Use the SDK when running outside Workers or when you need OpenFeature vendor-neutrality.

## Reading Order

| Task | Read |
|------|------|
| Set up Flagship in a Worker | `configuration.md` → `api.md` |
| Evaluate flags in code | `configuration.md` → `patterns.md` |
| Manage flags via REST API | `api.md` → `patterns.md` |
| Design targeting rules & rollouts | `patterns.md` → `gotchas.md` |
| Debug flag evaluation issues | `gotchas.md` → `api.md` |

REST API note: management endpoints use Cloudflare v4 envelopes (`result`, `result_info`, `errors`) and snake_case fields. The `/evaluate` endpoint is the exception: it is not enveloped and returns OpenFeature-style camelCase.

## In This Reference

- **[api.md](./api.md)** — REST API endpoints, binding methods, OpenFeature SDK, schemas
- **[configuration.md](./configuration.md)** — Wrangler binding setup, SDK installation, TypeScript types
- **[patterns.md](./patterns.md)** — Flag CRUD via API, targeting rules, rollouts, OpenFeature usage
- **[gotchas.md](./gotchas.md)** — Common errors, limits, anti-patterns, troubleshooting

## See Also

- **[Flagship API reference](https://developers.cloudflare.com/api/resources/flagship/)** — Source of truth for REST API paths, envelopes, and response fields
- **[Workers docs](https://developers.cloudflare.com/workers/)** — Workers runtime (Flagship runs inside Workers)
- **[../kv/](../kv/)** — KV storage (Flagship uses KV infrastructure for flag delivery)
- **[Wrangler docs](https://developers.cloudflare.com/workers/wrangler/)** — Wrangler CLI for deployment and config

# GraphQL Analytics API Reference

## Query Root

The schema has a single entry point: `Query.viewer`. Mutations are not supported.

```graphql
{
  cost       # uint64 -- query cost (returned in response)
  viewer {
    budget   # uint64 -- remaining budget
    zones(filter: { zoneTag: "..." }) { ... }
    accounts(filter: { accountTag: "..." }) { ... }
  }
}
```

## Aggregation Fields

Aggregated dataset nodes (`*Groups`) return these field categories. Not every node has all — use introspection to check.

### count

Total events in the group. Available on `*Groups` nodes but **not** on raw `*Adaptive` nodes (e.g., `workersInvocationsAdaptive` — use `sum { requests }` instead).

### sum

Cumulative metrics. Fields vary by dataset:

```graphql
# HTTP requests
sum { edgeResponseBytes edgeRequestBytes visits edgeTimeToFirstByteMs originResponseDurationMs }

# Workers invocations
sum { requests errors subrequests cpuTimeUs wallTime duration responseBodySize clientDisconnects requestDuration }
```

### quantiles

Percentile distributions (on datasets like `workersInvocationsAdaptive`). Available percentiles: P25, P50, P75, P90, P95, P99, P999 for `cpuTime`, `wallTime`, `requestDuration`, `duration`, `responseBodySize`.

```graphql
quantiles { cpuTimeP50 cpuTimeP99 wallTimeP50 wallTimeP99 }
```

### ratio, avg, uniq, confidence

```graphql
ratio { status4xx status5xx }      # float64 (0 to 1) -- HTTP datasets only
avg { sampleInterval }              # useful for understanding sampling resolution
uniq { uniques }                    # unique IP count -- rollup datasets (*1hGroups, *1dGroups) only
confidence(level: 0.95) {           # Adaptive datasets only; works on count and sum fields
  count { estimate lower upper sampleSize }
}
```

## Dimensions

Dimensions are fields you can group by via the `dimensions` sub-selection.

### Time Dimensions

| Dimension | Granularity |
|-----------|------------|
| `date` | Day |
| `datetime` | Exact timestamp |
| `datetimeMinute` | 1 minute |
| `datetimeFiveMinutes` | 5 minutes |
| `datetimeFifteenMinutes` | 15 minutes |
| `datetimeHour` | 1 hour |

Workers datasets also support `datetimeSixHours`.

### HTTP Request Dimensions (httpRequestsAdaptiveGroups)

83 dimensions available. Key ones:

| Dimension | Description |
|-----------|-------------|
| `clientCountryName` | Country of origin |
| `clientRequestHTTPHost` | Requested hostname |
| `clientRequestHTTPMethodName` | HTTP method |
| `clientRequestPath` | URI path |
| `edgeResponseStatus` | Edge HTTP status code |
| `cacheStatus` | Cache status (hit, miss, dynamic, etc.) |
| `coloCode` | Cloudflare datacenter IATA code |
| `clientIP` / `clientAsn` | Client IP address / ASN |
| `botScore` / `botManagementDecision` | Bot management score (0-99) / verdict |
| `wafAttackScore` / `securityAction` | WAF score / firewall action taken |
| `ja3Hash` / `ja4` | TLS fingerprints |
| `sampleInterval` | ABR sample interval |

### Workers Dimensions (workersInvocationsAdaptive)

`scriptName`, `scriptTag`, `scriptVersion`, `environmentName`, `status`, `usageModel`, `coloCode`, `dispatchNamespaceName`, `isDispatcher`

### Firewall Dimensions (firewallEventsAdaptive)

`action`, `source`, `ruleId`, `clientCountryName`, `clientIP`, `clientAsn`, `userAgent`

## Filtering

### Scope Filters

```graphql
zones(filter: { zoneTag: "ZONE_ID" })           # up to 10 zones
zones(filter: { zoneTag_in: ["Z1", "Z2"] })
accounts(filter: { accountTag: "ACCOUNT_ID" })   # exactly 1 account
```

### Dataset Filters

**Always include a time range filter.** Multiple filters at the same level are implicitly AND-ed.

```graphql
httpRequestsAdaptiveGroups(
  filter: { datetime_gt: "2025-01-01T00:00:00Z", datetime_lt: "2025-01-02T00:00:00Z", clientCountryName: "US" }
  limit: 1000
)
```

### Filter Operators

| Operator | Meaning | Example |
|----------|---------|---------|
| (none) | equals | `clientCountryName: "US"` |
| `_gt` / `_lt` | greater / less than | `datetime_gt: "..."` |
| `_geq` / `_leq` | greater/less or equal | `datetime_geq: "..."` |
| `_neq` | not equal | `cacheStatus_neq: "hit"` |
| `_in` / `_notin` | in / not in list | `clientCountryName_in: ["US", "GB"]` |
| `_like` / `_notlike` | SQL LIKE with `%` | `clientRequestPath_like: "/api/%"` |
| `_has` / `_hasall` / `_hasany` | array contains | `botDetectionIds_has: "abc"` |

> `_notin` and `_notlike` are in the schema but not in official docs. Confirmed via introspection.

### Boolean Operators (AND / OR)

```graphql
# Explicit AND
filter: { AND: [{ datetime_gt: "..." }, { datetime_lt: "..." }, { clientCountryName: "US" }] }

# Explicit OR
filter: { datetime_gt: "...", OR: [{ edgeResponseStatus: 403 }, { edgeResponseStatus: 429 }] }
```

## Pagination & Sorting

No cursor-based pagination. Use `limit`, `orderBy`, and filter-based offsets:

```graphql
# First page
httpRequestsAdaptiveGroups(filter: { datetime_gt: "..." }, limit: 100, orderBy: [datetime_ASC])

# Next page: filter by last seen value from previous page
httpRequestsAdaptiveGroups(filter: { datetime_gt: "2025-01-01T01:35:00Z" }, limit: 100, orderBy: [datetime_ASC])
```

Sort with `orderBy: [field_ASC]` or `[field_DESC]`. Multiple sort fields supported.

## Settings Node

Query per-node limits and availability:

```graphql
viewer { zones(filter: { zoneTag: "..." }) { settings {
  httpRequestsAdaptiveGroups { enabled maxDuration maxNumberOfFields maxPageSize notOlderThan }
} } }
```

## See Also

- [README.md](README.md) - Overview, decision tree, dataset index
- [configuration.md](configuration.md) - Authentication, client setup, introspection queries
- [patterns.md](patterns.md) - Common query patterns (time-series, top-N, per-product)
- [gotchas.md](gotchas.md) - Rate limits, sampling, troubleshooting

# GraphQL Analytics API Configuration

## Authentication

### API Token (Recommended)

| Permission | Scope | Use Case |
|------------|-------|----------|
| **Account Analytics: Read** | Account-wide | Workers, R2, KV, D1, DO, AI, Network Analytics |
| **Zone Analytics: Read** | Per-zone | HTTP requests, Firewall, DNS, Load Balancing |
| **All zones - Analytics: Read** | All zones | Multi-zone HTTP/Firewall/DNS queries |

Create tokens at: [dash.cloudflare.com > Account API Tokens](https://dash.cloudflare.com/?to=/:account/api-tokens)

```bash
# Verify token
curl -s https://api.cloudflare.com/client/v4/graphql \
  -H "Authorization: Bearer $CF_API_TOKEN" \
  -H "Content-Type: application/json" \
  --data '{"query":"{ viewer { zones(filter: {zoneTag: \"ZONE_ID\"}) { httpRequestsAdaptiveGroups(limit: 1, filter: {datetime_gt: \"2025-01-01T00:00:00Z\"}) { count } } } }"}'
```

### API Key + Email (Legacy)

Not recommended. Use `X-Auth-Email` + `X-Auth-Key` headers instead of `Authorization: Bearer`.

## Client Setup

### curl

```bash
curl -s https://api.cloudflare.com/client/v4/graphql \
  -H "Authorization: Bearer $CF_API_TOKEN" \
  -H "Content-Type: application/json" \
  --data '{
    "query": "query($zoneTag: string!, $start: Time!, $end: Time!) { viewer { zones(filter: {zoneTag: $zoneTag}) { httpRequestsAdaptiveGroups(filter: {datetime_gt: $start, datetime_lt: $end}, limit: 10, orderBy: [datetimeFiveMinutes_DESC]) { count dimensions { datetimeFiveMinutes } } } } }",
    "variables": { "zoneTag": "ZONE_ID", "start": "2025-01-01T00:00:00Z", "end": "2025-01-02T00:00:00Z" }
  }' | jq .
```

### TypeScript / JavaScript

```typescript
const GRAPHQL_ENDPOINT = "https://api.cloudflare.com/client/v4/graphql";

async function queryGraphQL<T>(query: string, variables: Record<string, unknown> = {}): Promise<T> {
  const response = await fetch(GRAPHQL_ENDPOINT, {
    method: "POST",
    headers: { Authorization: `Bearer ${process.env.CF_API_TOKEN}`, "Content-Type": "application/json" },
    body: JSON.stringify({ query, variables }),
  });
  if (!response.ok) throw new Error(`HTTP ${response.status}: ${response.statusText}`);
  const json = await response.json() as { data: T | null; errors?: { message: string }[] };
  if (json.errors?.length) throw new Error(json.errors.map((e) => e.message).join("; "));
  return json.data!;
}
```

### Python

```python
import requests, os

def query_graphql(query: str, variables: dict = None) -> dict:
    r = requests.post("https://api.cloudflare.com/client/v4/graphql",
        headers={"Authorization": f"Bearer {os.environ['CF_API_TOKEN']}", "Content-Type": "application/json"},
        json={"query": query, "variables": variables or {}})
    r.raise_for_status()
    result = r.json()
    if result.get("errors"):
        raise Exception("; ".join(e["message"] for e in result["errors"]))
    return result["data"]
```

### From a Cloudflare Worker

Store the API token as a secret (`CF_API_TOKEN`). Use standard `fetch` to POST to `https://api.cloudflare.com/client/v4/graphql` with the same JSON body format as above. Always check `response.errors` — GraphQL returns 200 even on query failures.

## GraphQL API Explorer

Interactive explorer at [graphql.cloudflare.com](https://graphql.cloudflare.com/) — provides schema docs, autocomplete, variable panel, and shareable queries. Authenticates via your Cloudflare dashboard session.

## Schema Introspection

```graphql
# List zone-scoped datasets
{ __type(name: "zone") { fields { name description } } }

# List account-scoped datasets
{ __type(name: "account") { fields { name description } } }

# Discover dimensions for a dataset
{ __type(name: "ZoneHttpRequestsAdaptiveGroupsDimensions") {
  fields { name type { name kind } }
} }

# Discover filter operators for a dataset
{ __type(name: "ZoneHttpRequestsAdaptiveGroupsFilter_InputObject") {
  inputFields { name type { name kind } }
} }
```

## Finding Your Zone and Account IDs

- **Zone ID**: Dashboard > select zone > Overview (right sidebar), or via API
- **Account ID**: Dashboard > Account Home URL, or via API

```bash
curl -s https://api.cloudflare.com/client/v4/zones -H "Authorization: Bearer $CF_API_TOKEN" | jq '.result[] | {name, id}'
curl -s https://api.cloudflare.com/client/v4/accounts -H "Authorization: Bearer $CF_API_TOKEN" | jq '.result[] | {name, id}'
```

## See Also

- [README.md](README.md) - Overview, decision tree, dataset index
- [api.md](api.md) - Query structure, aggregation fields, filtering operators
- [patterns.md](patterns.md) - Common query patterns (time-series, top-N, per-product)
- [gotchas.md](gotchas.md) - Rate limits, sampling, troubleshooting

# GraphQL Analytics API Gotchas & Troubleshooting

## Rate Limits

| Limit | Value |
|-------|-------|
| GraphQL queries per user | **Default 300 per 5 minutes** (max 320, at least 1/sec) |
| General API rate limit | 1200 requests per 5 minutes (shared across all API calls) |
| Zone scope per query | Up to **10 zones** |
| Account scope per query | Exactly **1 account** |

The GraphQL rate limit is separate from the general API limit. Exceeding either results in `HTTP 429` and blocks all API calls for 5 minutes. Enterprise customers can contact support to raise limits.

### "429 Too Many Requests"

**Cause:** Exceeded rate limit.

**Solution:** Batch multiple datasets into single queries, cache results, increase intervals between queries. Use `{ viewer { budget } }` to monitor remaining budget.

## Sampling & Data Accuracy

### Adaptive Bit Rate (ABR) Sampling

Datasets with `Adaptive` in the name use adaptive sampling:
- Results are **statistically representative**, not exact
- Same query may return **slightly different numbers** each run
- Higher traffic = higher sampling rate = more accurate
- `sampleInterval` dimension shows the ratio (1 = no sampling, 10 = ~1-in-10 sampled)

For high-confidence numbers, use `confidence(level: 0.95)` to get estimate bounds. For exact counts, use rollup nodes (`httpRequests1hGroups`, `httpRequests1dGroups`) which are pre-aggregated without sampling.

### Rollup vs. Adaptive

| Feature | Rollup (`*1hGroups`, `*1dGroups`) | Adaptive (`*AdaptiveGroups`) |
|---------|-----------------------------------|-----------------------------|
| Sampling | No (pre-aggregated) | Yes (ABR) |
| Flexibility | Fixed time buckets | Any granularity |
| Dimensions | Fewer | Many more |
| Accuracy | Exact | Statistical estimate |

## Common Errors

### "Access denied" / "authentication error"

**Cause:** Token lacks required permission or wrong scope.

**Solution:** Account-scoped queries need **Account Analytics: Read**. Zone-scoped queries need **Zone Analytics: Read**. Verify: `curl -s https://api.cloudflare.com/client/v4/user/tokens/verify -H "Authorization: Bearer $TOKEN"`

### "field not found" / "Cannot query field"

**Cause:** Wrong dataset name, nonexistent field, or wrong scope (zone vs. account).

**Solution:** Names are case-sensitive camelCase (`httpRequestsAdaptiveGroups`). Zone datasets go under `zones(...)`, account datasets under `accounts(...)`. Use introspection to verify.

### "filter is required" / empty results

**Cause:** Missing required time range filter or incorrect zone/account tag.

**Solution:** Always include `datetime_gt` / `datetime_lt` (or `_geq` / `_leq`).

### "limit is required" / "limit exceeds maximum"

**Cause:** Missing `limit` or exceeding node's max page size.

**Solution:** Always specify `limit`. Max varies by dataset (typically 10,000 for groups, 100 for raw events). Check via settings query.

### "query is too complex" / "query exceeds budget"

**Cause:** Too many fields, datasets, or too broad a time range.

**Solution:** Reduce time range, request fewer dimensions/metrics, break into smaller queries. Monitor `cost` and `budget` in responses.

### 200 Response with Errors

GraphQL returns HTTP 200 even on failures. **Always check `response.errors`:**

```json
{ "data": null, "errors": [{ "message": "filter is required for httpRequestsAdaptiveGroups" }] }
```

## Plan-Based Availability

Not all datasets are available on all plans. Higher plans get more datasets, longer retention (`notOlderThan`), wider time ranges (`maxDuration`), more fields, and larger page sizes.

### "node is not available" / "node is disabled"

**Cause:** Dataset not on your plan, or product not enabled.

**Solution:** Check `settings { <nodeName> { enabled } }`. Some datasets require specific subscriptions (e.g., Network Analytics requires Magic Transit/Spectrum).

## DateTime & Timezone Handling

- All times are **UTC only** (ISO 8601: `"2025-01-15T10:30:00Z"`)
- `Date` type: `"2025-01-15"` (used in `date_geq`/`date_leq` for storage datasets)
- `Time` type: `"2025-01-15T10:30:00Z"` (used in `datetime_gt`/`datetime_lt`)
- Filters are start-inclusive: events that start within the window are included

## Performance Tips

- **Narrow time ranges** are faster and cheaper
- **Select only needed dimensions** — each additional dimension increases cost
- **Use rollup nodes** (`*1dGroups`) for simple daily totals without dimension breakdowns
- **Batch datasets** into one query instead of separate HTTP requests

## See Also

- [README.md](README.md) - Overview, decision tree, dataset index
- [api.md](api.md) - Query structure, aggregation fields, filtering operators
- [configuration.md](configuration.md) - Authentication, client setup, introspection queries
- [patterns.md](patterns.md) - Common query patterns (time-series, top-N, per-product)

# GraphQL Analytics API Patterns & Best Practices

## Time-Series Queries

Use time dimension granularity matching your range (see Best Practices below).

```graphql
query TrafficTimeSeries($zoneTag: string!, $start: Time!, $end: Time!) {
  viewer {
    zones(filter: { zoneTag: $zoneTag }) {
      httpRequestsAdaptiveGroups(
        filter: { datetime_gt: $start, datetime_lt: $end }
        limit: 1000
        orderBy: [datetimeFiveMinutes_ASC]  # or datetimeHour_ASC for longer ranges
      ) {
        count
        dimensions { datetimeFiveMinutes }
        sum { edgeResponseBytes }
        ratio { status4xx status5xx }
      }
    }
  }
}
```

## Top-N Queries

### Top Countries by Request Count

```graphql
query TopCountries($zoneTag: string!, $start: Time!, $end: Time!) {
  viewer {
    zones(filter: { zoneTag: $zoneTag }) {
      httpRequestsAdaptiveGroups(
        filter: { datetime_gt: $start, datetime_lt: $end }
        limit: 10
        orderBy: [count_DESC]
      ) {
        count
        dimensions { clientCountryName }
      }
    }
  }
}
```

Use `orderBy: [sum_edgeResponseBytes_DESC]` for top paths by bandwidth. Add `edgeResponseStatus_geq: 400` to the filter for top error status codes.

## Workers Analytics

```graphql
query WorkersOverview($accountTag: string!, $start: Time!, $end: Time!) {
  viewer {
    accounts(filter: { accountTag: $accountTag }) {
      workersInvocationsAdaptive(
        filter: { datetime_gt: $start, datetime_lt: $end }
        limit: 100
        orderBy: [sum_requests_DESC]
      ) {
        sum { requests errors subrequests wallTime }
        quantiles { cpuTimeP50 cpuTimeP99 wallTimeP50 wallTimeP99 }
        dimensions { scriptName }
      }
    }
  }
}
```

Filter by `scriptName` for a specific Worker. Add `datetimeFiveMinutes` dimension + `orderBy: [datetimeFiveMinutes_ASC]` for error rate over time.

## Firewall / Security

```graphql
query RecentFirewallEvents($zoneTag: string!, $start: Time!) {
  viewer {
    zones(filter: { zoneTag: $zoneTag }) {
      firewallEventsAdaptive(
        filter: { datetime_gt: $start }
        limit: 50
        orderBy: [datetime_DESC]
      ) {
        action source clientIP clientCountryName userAgent
        clientRequestHTTPHost clientRequestPath ruleId datetime
      }
    }
  }
}
```

For aggregated firewall stats, use `firewallEventsAdaptiveGroups` with `action: "block"` filter and group by `ruleId`, `source`, `datetimeHour`.

## DNS Analytics

```graphql
query DNSQueryVolume($zoneTag: string!, $start: Time!, $end: Time!) {
  viewer {
    zones(filter: { zoneTag: $zoneTag }) {
      dnsAnalyticsAdaptiveGroups(
        filter: { datetime_gt: $start, datetime_lt: $end }
        limit: 500
        orderBy: [datetimeFiveMinutes_ASC]
      ) {
        count
        dimensions { datetimeFiveMinutes }
      }
    }
  }
}
```

## Storage Analytics (Account-Scoped)

R2, KV, and D1 use `date` (Date type) filters instead of `datetime` (Time type).

```graphql
# R2 operations
r2OperationsAdaptiveGroups(filter: { date_geq: $start, date_leq: $end }, limit: 100, orderBy: [date_DESC]) {
  dimensions { date bucketName actionType }
  sum { requests }
}

# KV operations
kvOperationsAdaptiveGroups(filter: { date_geq: $start, date_leq: $end }, limit: 100, orderBy: [date_DESC]) {
  dimensions { date actionType }
  sum { requests }
}

# D1 analytics
d1AnalyticsAdaptiveGroups(filter: { date_geq: $start, date_leq: $end }, limit: 100, orderBy: [date_DESC]) {
  dimensions { date databaseId }
  sum { readQueries writeQueries rowsRead rowsWritten }
}
```

## Cache Analytics

```graphql
query CacheStatusBreakdown($zoneTag: string!, $start: Time!, $end: Time!) {
  viewer {
    zones(filter: { zoneTag: $zoneTag }) {
      httpRequestsAdaptiveGroups(
        filter: { datetime_gt: $start, datetime_lt: $end }
        limit: 20
        orderBy: [count_DESC]
      ) {
        count
        dimensions { cacheStatus }
        sum { edgeResponseBytes }
      }
    }
  }
}
```

For cache hit ratio over time, use aliases to query the same dataset twice — once with `cacheStatus: "hit"` filter and once without — then compute the ratio client-side.

## Multi-Dataset Queries

A single request can query multiple datasets, avoiding extra HTTP round-trips:

```graphql
query DashboardOverview($zoneTag: string!, $start: Time!, $end: Time!) {
  viewer {
    zones(filter: { zoneTag: $zoneTag }) {
      httpTraffic: httpRequestsAdaptiveGroups(
        filter: { datetime_gt: $start, datetime_lt: $end }, limit: 1
      ) { count  sum { edgeResponseBytes }  ratio { status4xx status5xx } }
      firewallEvents: firewallEventsAdaptiveGroups(
        filter: { datetime_gt: $start, datetime_lt: $end }, limit: 5, orderBy: [count_DESC]
      ) { count  dimensions { action source } }
      dnsQueries: dnsAnalyticsAdaptiveGroups(
        filter: { datetime_gt: $start, datetime_lt: $end }, limit: 1
      ) { count }
    }
  }
}
```

## AI & Gateway Analytics

```graphql
# Workers AI inference
aiInferenceAdaptiveGroups(
  filter: { datetime_gt: $start, datetime_lt: $end }, limit: 100, orderBy: [datetimeHour_DESC]
) {
  count
  sum { totalInputTokens totalOutputTokens totalRequestBytesIn }
  dimensions { modelId datetimeHour }
}

# AI Gateway requests
aiGatewayRequestsAdaptiveGroups(
  filter: { datetime_gt: $start, datetime_lt: $end }, limit: 100, orderBy: [datetimeHour_DESC]
) {
  count
  dimensions { gateway provider model datetimeHour }
  sum { cachedTokensIn cachedTokensOut uncachedTokensIn uncachedTokensOut }
}
```

Both are account-scoped — nest under `accounts(filter: { accountTag: $accountTag })`.

## Best Practices

**Always include time filters.** Queries without time filters scan all data and are slow/expensive.

**Match time granularity to range:**

| Time Range | Recommended Dimension |
|------------|----------------------|
| < 6 hours | `datetimeMinute` or `datetimeFiveMinutes` |
| 6-48 hours | `datetimeFiveMinutes` or `datetimeFifteenMinutes` |
| 2-14 days | `datetimeHour` |
| 14+ days | `date` |

**Use aliases** for querying the same dataset with different filters in one request.

**Request only needed fields.** Extra dimensions and metrics increase query cost.

## See Also

- [README.md](README.md) - Overview, decision tree, dataset index
- [api.md](api.md) - Query structure, aggregation fields, filtering operators
- [configuration.md](configuration.md) - Authentication, client setup, introspection queries
- [gotchas.md](gotchas.md) - Rate limits, sampling, troubleshooting

# Cloudflare GraphQL Analytics API

Query analytics data across all Cloudflare products via a single GraphQL endpoint. Covers HTTP requests, Workers metrics, DNS, Firewall events, Network Analytics, and 70+ other datasets.

## Overview

- **Single endpoint** for all analytics: `https://api.cloudflare.com/client/v4/graphql`
- **1,400+ schema types** spanning every Cloudflare product
- **Two scopes**: zone-level (per-domain) and account-level (cross-domain)
- **Adaptive sampling** on high-traffic datasets with confidence intervals
- **No mutations** - read-only analytics (the Mutation type is a stub)
- **Cost-based rate limiting** - default 300 queries per 5 minutes per user (max 320, varies by query cost)

## Quick Decision Tree

```
Need analytics data from Cloudflare?
├─ HTTP traffic (requests, bandwidth, cache) → httpRequestsAdaptiveGroups (zone or account)
├─ Workers performance (CPU, wall time, errors) → workersInvocationsAdaptive (account)
├─ Firewall/WAF events → firewallEventsAdaptive / firewallEventsAdaptiveGroups (zone or account)
├─ DNS query analytics → dnsAnalyticsAdaptive / dnsAnalyticsAdaptiveGroups (zone or account)
├─ Network layer (DDoS, Magic Transit) → *NetworkAnalyticsAdaptiveGroups (account)
├─ Storage (R2, KV, D1, DO) → r2OperationsAdaptiveGroups / kvOperationsAdaptiveGroups / etc. (account)
├─ AI (Workers AI, AI Gateway) → aiInferenceAdaptive / aiGatewayRequestsAdaptiveGroups (account)
├─ Load Balancing → loadBalancingRequestsAdaptiveGroups (zone)
├─ Custom high-cardinality metrics → Workers Analytics Engine (see ../analytics-engine/)
└─ Need raw logs, not aggregates → Logpush (see Cloudflare docs)
```

## Core Concepts

| Concept | Description |
|---------|-------------|
| **Endpoint** | `POST https://api.cloudflare.com/client/v4/graphql` |
| **Explorer** | [graphql.cloudflare.com](https://graphql.cloudflare.com/) - interactive query builder |
| **Viewer** | Root query object: `viewer { zones(...) { ... } }` or `viewer { accounts(...) { ... } }` |
| **Dataset (Node)** | A queryable table under a zone or account (e.g., `httpRequestsAdaptiveGroups`) |
| **Dimensions** | Fields to group by (time buckets, country, status code, script name, etc.) |
| **Metrics** | Aggregation fields: `count`, `sum { ... }`, `avg { ... }`, `quantiles { ... }`, `ratio { ... }` |
| **Filter** | Input object constraining results by time range, dimensions, etc. |
| **Limit** | Maximum rows returned per dataset node (required, max varies by dataset) |
| **OrderBy** | Enum-based sorting: `[field_ASC]` or `[field_DESC]` |
| **Adaptive Sampling** | Nodes with `Adaptive` in the name use ABR sampling; results are statistically representative |

## Query Structure

Every query follows this pattern:

```graphql
{
  viewer {
    # Zone-scoped
    zones(filter: { zoneTag: "ZONE_ID" }) {
      datasetName(
        filter: { datetime_gt: "...", datetime_lt: "..." }
        limit: 1000
        orderBy: [datetimeFiveMinutes_DESC]
      ) {
        count
        dimensions { ... }
        sum { ... }
      }
    }
    # Account-scoped
    accounts(filter: { accountTag: "ACCOUNT_ID" }) {
      datasetName(filter: { ... }, limit: 100) {
        count
        dimensions { ... }
        sum { ... }
      }
    }
  }
}
```

## Dataset Naming Convention

Dataset names follow a consistent pattern visible in the schema:

| Pattern | Meaning | Example |
|---------|---------|---------|
| `*Adaptive` | Raw rows with adaptive sampling; some (e.g., `workersInvocationsAdaptive`) also support aggregation fields (`sum`, `quantiles`, `avg`) | `httpRequestsAdaptive`, `workersInvocationsAdaptive` |
| `*AdaptiveGroups` | Aggregated data with adaptive sampling | `httpRequestsAdaptiveGroups` |
| `*1hGroups` | Hourly rollups (pre-aggregated) | `httpRequests1hGroups` |
| `*1dGroups` | Daily rollups (pre-aggregated) | `httpRequests1dGroups` |
| `*1mGroups` | Minutely rollups | `httpRequests1mGroups` |
| `Zone*` prefix | Zone-scoped dataset | `ZoneHttpRequestsAdaptiveGroups` |
| `Account*` prefix | Account-scoped dataset | `AccountWorkersInvocationsAdaptive` |

**Prefer `*AdaptiveGroups` nodes** for most use cases - they support flexible time grouping via dimension fields (`datetimeFiveMinutes`, `datetimeHour`, etc.) and are the most commonly used.

## Key Datasets by Product

### Zone-Scoped (per-domain)

| Dataset | Description |
|---------|-------------|
| `httpRequestsAdaptiveGroups` | HTTP traffic: requests, bytes, cache status, bot scores, WAF scores |
| `httpRequests1hGroups` / `1dGroups` / `1mGroups` | Pre-aggregated HTTP rollups (hourly/daily/minutely) |
| `firewallEventsAdaptiveGroups` | WAF, rate limiting, bot management, firewall rule events |
| `dnsAnalyticsAdaptiveGroups` | DNS query volumes, response codes, query types |
| `loadBalancingRequestsAdaptiveGroups` | Load Balancer origin request metrics |
| `pageShieldReportsAdaptiveGroups` | Page Shield CSP reports |

### Account-Scoped (cross-domain)

| Dataset | Description |
|---------|-------------|
| `workersInvocationsAdaptive` | Workers: requests, errors, CPU time, wall time, subrequests |
| `durableObjectsInvocationsAdaptiveGroups` | DO invocations |
| `durableObjectsStorageGroups` / `durableObjectsPeriodicGroups` | DO storage and periodic metrics |
| `d1AnalyticsAdaptiveGroups` / `d1QueriesAdaptiveGroups` | D1 database analytics |
| `r2OperationsAdaptiveGroups` / `r2StorageAdaptiveGroups` | R2 operations and storage |
| `kvOperationsAdaptiveGroups` / `kvStorageAdaptiveGroups` | KV operations and storage |
| `aiInferenceAdaptiveGroups` | Workers AI inference metrics |
| `aiGatewayRequestsAdaptiveGroups` | AI Gateway request analytics |
| `pagesFunctionsInvocationsAdaptiveGroups` | Pages Functions metrics |
| `magicTransitNetworkAnalyticsAdaptiveGroups` | Magic Transit packet/byte analytics |
| `spectrumNetworkAnalyticsAdaptiveGroups` | Spectrum TCP/UDP analytics |
| `gatewayL7RequestsAdaptiveGroups` | Zero Trust Gateway HTTP metrics |
| `gatewayResolverQueriesAdaptiveGroups` | Zero Trust Gateway DNS metrics |

## Reading Order

| Task | Start Here | Then Read |
|------|------------|-----------|
| **First query** | [configuration.md](configuration.md) (auth) -> this README (structure) | [api.md](api.md) |
| **Build a dashboard** | [patterns.md](patterns.md) (time-series, top-N) | [api.md](api.md) (aggregation fields) |
| **Debug query issues** | [gotchas.md](gotchas.md) | [api.md](api.md) (filtering) |
| **Understand sampling** | [gotchas.md](gotchas.md) (sampling section) | [api.md](api.md) (confidence intervals) |
| **Product-specific metrics** | [patterns.md](patterns.md) (per-product examples) | [api.md](api.md) (dataset reference) |

## In This Reference

- **[api.md](api.md)** - Query structure, aggregation fields (sum/avg/quantiles/count), filtering operators, dimensions, dataset details
- **[configuration.md](configuration.md)** - Authentication, API tokens, client setup (curl, JS, Python), introspection
- **[patterns.md](patterns.md)** - Common queries: time-series, top-N, Workers metrics, HTTP analytics, firewall events, multi-zone
- **[gotchas.md](gotchas.md)** - Rate limits, sampling caveats, query cost, common errors, plan-based limits

## See Also

- [GraphQL Analytics API Docs](https://developers.cloudflare.com/analytics/graphql-api/)
- [GraphQL API Explorer](https://graphql.cloudflare.com/)
- [Observability Reference](../observability/) - Workers Logs, Tail Workers, console logging
- [Analytics Engine Reference](../analytics-engine/) - Custom high-cardinality analytics via Workers
- [Web Analytics Reference](../web-analytics/) - Client-side (RUM) analytics
- [API Reference](../api/) - REST API, SDKs, authentication basics

# Hyperdrive API and drivers

Start with [README.md](./README.md) and [configuration.md](./configuration.md). Fetch the selected guide before writing connection or query code; use its current supported package version and compatibility settings.

## Driver and binding routes

| Task | Official documentation |
|------|------------------------|
| PostgreSQL with node-postgres (`pg`), including binding connection string and parameterized queries | [node-postgres](https://developers.cloudflare.com/hyperdrive/examples/connect-to-postgres/postgres-drivers-and-libraries/node-postgres/) |
| PostgreSQL with tagged-template queries and Postgres.js driver options | [Postgres.js](https://developers.cloudflare.com/hyperdrive/examples/connect-to-postgres/postgres-drivers-and-libraries/postgres-js/) |
| MySQL with binding connection properties and Worker-specific driver options | [mysql2](https://developers.cloudflare.com/hyperdrive/examples/connect-to-mysql/mysql-drivers-and-libraries/mysql2/) |
| Check database features, prepared statements, and library compatibility | [Supported databases and features](https://developers.cloudflare.com/hyperdrive/reference/supported-databases-and-features/) |
| Generate binding and runtime TypeScript types | [Workers TypeScript](https://developers.cloudflare.com/workers/languages/typescript/) |

Keep an existing supported driver when it fits the application. Choose by database engine and library integration needs; do not infer cache behavior from a driver's prepared-statement setting. Fetch [query caching](https://developers.cloudflare.com/hyperdrive/concepts/query-caching/) for cache eligibility and freshness controls.

## ORMs and query builders

| Task | Official documentation |
|------|------------------------|
| Use Drizzle with PostgreSQL | [PostgreSQL Drizzle guide](https://developers.cloudflare.com/hyperdrive/examples/connect-to-postgres/postgres-drivers-and-libraries/drizzle-orm/) |
| Use Drizzle with MySQL | [MySQL Drizzle guide](https://developers.cloudflare.com/hyperdrive/examples/connect-to-mysql/mysql-drivers-and-libraries/drizzle-orm/) |
| Use Prisma with PostgreSQL | [Prisma guide](https://developers.cloudflare.com/hyperdrive/examples/connect-to-postgres/postgres-drivers-and-libraries/prisma-orm/) |
| Assess another query builder, including Kysely | [Postgres.js integration notes](https://developers.cloudflare.com/hyperdrive/examples/connect-to-postgres/postgres-drivers-and-libraries/postgres-js/) and [database compatibility](https://developers.cloudflare.com/hyperdrive/reference/supported-databases-and-features/), then the library's current dialect documentation |

An ORM still uses a database driver and inherits its Worker connection constraints. Keep clients scoped to the invocation using [connection lifecycle](https://developers.cloudflare.com/hyperdrive/concepts/connection-lifecycle/). When a library owns SQL for authentication or other fresh reads, pass a client using a cache-disabled configuration as described in [query caching](https://developers.cloudflare.com/hyperdrive/concepts/query-caching/).

# Hyperdrive configuration

See [README.md](./README.md) for the retrieval workflow. Fetch the relevant guide before creating or changing resources; use current configuration fields and CLI syntax from these sources.

| Task | Official documentation |
|------|------------------------|
| Create the first configuration and bind it to a Worker | [Get started](https://developers.cloudflare.com/hyperdrive/get-started/) |
| Create, inspect, update, or delete configurations; set cache or pool options | [Wrangler commands](https://developers.cloudflare.com/hyperdrive/reference/wrangler-commands/) |
| Generate TypeScript types from Worker configuration | [Workers TypeScript](https://developers.cloudflare.com/workers/languages/typescript/) |
| Connect a private database using the recommended Workers VPC route | [Workers VPC integration](https://developers.cloudflare.com/hyperdrive/configuration/connect-to-private-database-vpc/) |
| Maintain a private database connection using Tunnel and Access | [Tunnel integration](https://developers.cloudflare.com/hyperdrive/configuration/connect-to-private-database/) |
| Configure database network access | [Firewall and networking](https://developers.cloudflare.com/hyperdrive/configuration/firewall-and-networking-configuration/) |
| Configure server verification or client certificates | [SSL/TLS certificates](https://developers.cloudflare.com/hyperdrive/configuration/tls-ssl-certificates-for-hyperdrive/) |
| Rotate origin database credentials | [Credential rotation](https://developers.cloudflare.com/hyperdrive/configuration/rotate-credentials/) |
| Configure cache freshness or separate cached and fresh-read bindings | [Query caching](https://developers.cloudflare.com/hyperdrive/concepts/query-caching/) |
| Budget origin connections across configurations | [Tune connection pooling](https://developers.cloudflare.com/hyperdrive/configuration/tune-connection-pool/) |
| Choose local database access or remote Hyperdrive testing | [Local development](https://developers.cloudflare.com/hyperdrive/configuration/local-development/) |
| Evaluate Worker placement for multiple database round trips | [Smart Placement](https://developers.cloudflare.com/workers/configuration/placement/) |

## Setup decisions

- Identify the database engine, provider, and network path first. The [PostgreSQL](https://developers.cloudflare.com/hyperdrive/examples/connect-to-postgres/) and [MySQL](https://developers.cloudflare.com/hyperdrive/examples/connect-to-mysql/) indexes route to provider-specific instructions.
- For private connectivity, choose Workers VPC or the existing Tunnel/Access integration before configuring credentials. Follow the selected guide's prerequisites and TLS guidance.
- Decide which reads may be stale before selecting cache settings. Multiple configurations against one database contribute to its total origin connection usage.
- Local direct database access does not exercise Hyperdrive pooling or caching. Use the local-development guide's remote option when verifying those behaviors, and identify the database that option targets before running writes.

See [api.md](./api.md) for drivers and [gotchas.md](./gotchas.md) for diagnosis.

# Hyperdrive troubleshooting

Start with the actual error and the affected configuration. Fetch [Troubleshoot and debug](https://developers.cloudflare.com/hyperdrive/observability/troubleshooting/) for current error codes and diagnosis rather than guessing from a generic connection failure.

| Symptom | What to inspect and where to read |
|---------|----------------------------------|
| Connection refused or authentication failure | Check origin reachability and credentials using [troubleshooting](https://developers.cloudflare.com/hyperdrive/observability/troubleshooting/) and [firewall/networking configuration](https://developers.cloudflare.com/hyperdrive/configuration/firewall-and-networking-configuration/). |
| Private database or TLS failure | Follow the selected [Workers VPC](https://developers.cloudflare.com/hyperdrive/configuration/connect-to-private-database-vpc/) or [Tunnel/Access](https://developers.cloudflare.com/hyperdrive/configuration/connect-to-private-database/) path and its certificate prerequisites; see [SSL/TLS configuration](https://developers.cloudflare.com/hyperdrive/configuration/tls-ssl-certificates-for-hyperdrive/). |
| Pool exhaustion or too many connections | Distinguish client connection lifetime from origin pool capacity. Read [connection lifecycle](https://developers.cloudflare.com/hyperdrive/concepts/connection-lifecycle/), [pool tuning](https://developers.cloudflare.com/hyperdrive/configuration/tune-connection-pool/), and [limits](https://developers.cloudflare.com/hyperdrive/platform/limits/). |
| Query timeout | Check the current [limits](https://developers.cloudflare.com/hyperdrive/platform/limits/) and [metrics](https://developers.cloudflare.com/hyperdrive/observability/metrics/) before changing query or transaction design. |
| Stale reads or unexpectedly uncached queries | Inspect the binding's cache configuration and query eligibility in [query caching](https://developers.cloudflare.com/hyperdrive/concepts/query-caching/). Writes do not purge cached reads; do not treat prepared-statement settings as cache controls. |
| Slow multi-query requests | Inspect [metrics](https://developers.cloudflare.com/hyperdrive/observability/metrics/) and evaluate [Smart Placement](https://developers.cloudflare.com/workers/configuration/placement/). |
| Local connection failure, ignored environment variable, or absent cache behavior | Check binding names, local connection overrides, precedence, and remote testing in [local development](https://developers.cloudflare.com/hyperdrive/configuration/local-development/). |
| Unsupported driver or SQL feature | Check [supported databases and features](https://developers.cloudflare.com/hyperdrive/reference/supported-databases-and-features/) and the selected [driver guide](./api.md). |

## Capacity and changes

Retrieve [limits](https://developers.cloudflare.com/hyperdrive/platform/limits/) and [pricing](https://developers.cloudflare.com/hyperdrive/platform/pricing/) for current plan allowances, connection and query bounds, and limit-increase guidance. Check [release notes](https://developers.cloudflare.com/hyperdrive/platform/release-notes/) when behavior changes after an upgrade.

See [configuration.md](./configuration.md) to change a configuration and [patterns.md](./patterns.md) to revisit freshness or connection decisions.

# Hyperdrive design patterns

See [api.md](./api.md) for maintained driver and ORM examples. Use the following decisions to select a pattern, then fetch its linked documentation for implementation.

| Workload or decision | Guidance and documentation |
|----------------------|----------------------------|
| Popular content or analytics dashboards | Cache only when the product can tolerate the configured stale window. Use [query caching](https://developers.cloudflare.com/hyperdrive/concepts/query-caching/) for eligibility, parameters, and settings. |
| Mixed cached reads and fresh reads | Route authentication, permissions, and reads after writes through a cache-disabled configuration. Writes do not invalidate cached results; see [read-after-write behavior](https://developers.cloudflare.com/hyperdrive/concepts/query-caching/#read-after-write-behavior). |
| Multi-tenant queries | Derive tenant scope from authenticated application context and apply it to every query. A cache is not an authorization boundary. Review [query caching](https://developers.cloudflare.com/hyperdrive/concepts/query-caching/) for the selected query's behavior. |
| Globally distributed callers | Understand the distinction between fast connection setup and the remaining query round trip in [how Hyperdrive works](https://developers.cloudflare.com/hyperdrive/concepts/how-hyperdrive-works/). |
| Multiple sequential database queries | Measure placement rather than assuming the nearest user location is best. Consult [Smart Placement](https://developers.cloudflare.com/workers/configuration/placement/) and [Hyperdrive metrics](https://developers.cloudflare.com/hyperdrive/observability/metrics/). |
| Transactions or connection-local state | Keep transactions short and do not assume state survives across transactions. Fetch [connection pooling](https://developers.cloudflare.com/hyperdrive/concepts/connection-pooling/) and [supported features](https://developers.cloudflare.com/hyperdrive/reference/supported-databases-and-features/) before relying on session settings. |
| Client lifetime and pool sizing | Create clients per handler invocation; Hyperdrive owns the origin pool. Use [connection lifecycle](https://developers.cloudflare.com/hyperdrive/concepts/connection-lifecycle/) and [pool tuning](https://developers.cloudflare.com/hyperdrive/configuration/tune-connection-pool/) instead of a global driver pool or copied connection counts. |

Separate application correctness from acceleration: use parameterized queries, enforce tenant access in the application, and select freshness before tuning cache hit rate. See [gotchas.md](./gotchas.md) when observed behavior differs from the design.

# Hyperdrive

Use Hyperdrive to connect Workers to an existing PostgreSQL or MySQL database with connection pooling and optional query caching. It does not replace the origin database or replicate its data. Start with [how Hyperdrive works](https://developers.cloudflare.com/hyperdrive/concepts/how-hyperdrive-works/) and [supported databases and features](https://developers.cloudflare.com/hyperdrive/reference/supported-databases-and-features/) to assess fit.

## Retrieve current documentation

Fetch the relevant official page before implementing. Driver versions, compatibility settings, API shapes, CLI flags, cache settings, and limits belong in the docs rather than this reference. Use the [Hyperdrive documentation index](https://developers.cloudflare.com/hyperdrive/llms.txt) to discover additional pages. Retrieve a page as Markdown by sending `Accept: text/markdown` to its URL.

## Choose the next reference

| Task | Reference |
|------|-----------|
| Create a configuration, bind it, connect privately, or develop locally | [configuration.md](./configuration.md) |
| Choose a driver, use binding credentials, or integrate an ORM | [api.md](./api.md) |
| Decide read freshness, connection lifetime, or query placement | [patterns.md](./patterns.md) |
| Diagnose connection, cache, latency, or capacity problems | [gotchas.md](./gotchas.md) |

## Decisions to preserve

- Choose a driver for the database engine and existing application stack; verify supported versions and Worker requirements in its guide.
- Create database clients inside each handler invocation. Hyperdrive manages the underlying origin pool; consult [connection lifecycle](https://developers.cloudflare.com/hyperdrive/concepts/connection-lifecycle/) for cleanup behavior.
- Choose caching by read freshness. Disabling caching still allows connection pooling; a write does not invalidate cached reads. See [query caching](https://developers.cloudflare.com/hyperdrive/concepts/query-caching/).

## See also

- [D1](../d1/) for a managed SQLite alternative.
- [Workers](https://developers.cloudflare.com/workers/) for the runtime and bindings.

# Images API Selection

| Operation | Documentation |
|-----------|---------------|
| Optimize image bytes in a Worker; select input, transform, output, and response methods | [Optimize with Workers](https://developers.cloudflare.com/images/optimization/binding/#methods) |
| Upload, list, retrieve, update, or delete hosted images from a Worker | [Manage hosted images with Workers](https://developers.cloudflare.com/images/storage/binding/) |
| Upload or manage images through HTTP | [Upload methods](https://developers.cloudflare.com/images/storage/upload-images/methods/#upload-using-api) and its linked Images API reference |
| Accept uploads directly from a client | [Direct Creator Upload](https://developers.cloudflare.com/images/storage/upload-images/direct-creator-upload/) |
| Construct hosted-image delivery URLs | [Serve uploaded images](https://developers.cloudflare.com/images/optimization/hosted-images/serve-uploaded-images/) |
| Apply URL optimization parameters or select fit, quality, and format | [Optimization features](https://developers.cloudflare.com/images/optimization/features/) |
| Draw overlays or watermarks | [Draw overlays](https://developers.cloudflare.com/images/optimization/draw-overlays/) |

Do not transfer URL parameters or HTTP request shapes directly into binding calls. Read the contract for the selected interface, including output format handling. Use the project's generated binding types and existing error handling. See [configuration](configuration.md) for setup and [troubleshooting](gotchas.md) for failures and limits.

# Images Configuration

Inspect the project's Wrangler configuration, dependency versions, existing bindings, and credential storage before changing setup. Preserve its configuration format and generate binding types through its existing tooling.

| Task | Documentation |
|------|---------------|
| Add the optimization binding | [Binding setup](https://developers.cloudflare.com/images/optimization/binding/#setup) |
| Configure hosted-image management in a Worker | [Hosted binding setup](https://developers.cloudflare.com/images/storage/binding/#setup) |
| Choose local or remote development for the optimization binding | [Local binding development](https://developers.cloudflare.com/images/optimization/binding/#interact-with-your-images-binding-locally) |
| Upload through the dashboard or API | [Upload methods](https://developers.cloudflare.com/images/storage/upload-images/methods/) |
| Create named presets for hosted images | [Create predefined variants](https://developers.cloudflare.com/images/optimization/hosted-images/create-variants/) |
| Enable dynamic options for hosted-image URLs | [Enable flexible variants](https://developers.cloudflare.com/images/optimization/hosted-images/enable-flexible-variants/) |
| Find account hash and delivery URL components | [Serve uploaded images](https://developers.cloudflare.com/images/optimization/hosted-images/serve-uploaded-images/) |
| Configure private access and generate signed URLs | [Serve private images](https://developers.cloudflare.com/images/optimization/hosted-images/serve-private-images/) |
| Set hosted-image cache lifetime | [Browser TTL](https://developers.cloudflare.com/images/optimization/hosted-images/browser-ttl/) |

Keep API tokens and signing keys in the project's secret mechanism. Verify private-delivery requirements when choosing variants, and follow the documented signing procedure rather than maintaining a custom signing recipe here. Confirm that the selected local test mode covers the features being changed. Continue with [API selection](api.md) or [patterns](patterns.md).

# Images Troubleshooting

First identify whether the failure involves hosted-image storage, remote URL transformations, or a Worker binding. Capture the failing operation, response status, relevant headers, and error message before changing options.

| Symptom or question | Documentation |
|---------------------|---------------|
| Resizing is absent, an origin request fails, or a transformation returns an error code | [Troubleshooting](https://developers.cloudflare.com/images/reference/troubleshooting/) |
| Input size, dimensions, animation, or format compatibility | [Limits and formats](https://developers.cloudflare.com/images/get-started/limits/) — choose the section for the affected interface |
| Unexpected fit, quality, format, or crop behavior | [Optimization features](https://developers.cloudflare.com/images/optimization/features/) |
| Binding input, output, or response handling fails | [Binding methods](https://developers.cloudflare.com/images/optimization/binding/#methods) |
| Local behavior differs from production | [Local binding development](https://developers.cloudflare.com/images/optimization/binding/#interact-with-your-images-binding-locally) |
| Private delivery fails or an image is unexpectedly public | [Serve private images](https://developers.cloudflare.com/images/optimization/hosted-images/serve-private-images/) and [variant public access](https://developers.cloudflare.com/images/optimization/hosted-images/create-variants/#public-access) |
| Remote transformations appear stale | [Caching and purging](https://developers.cloudflare.com/images/reference/troubleshooting/#caching-and-purging) |
| Worker transformations repeat unnecessarily | [Binding caching guidance](https://developers.cloudflare.com/images/optimization/binding/#methods) |

Do not apply one interface's limits, error codes, or caching rules to another. Reproduce with a representative image and verify the chosen fix using the project's existing checks. Retry only after identifying a transient failure; changing invalid inputs or access configuration requires a different fix. See [API selection](api.md) and [configuration](configuration.md) when the wrong interface or setup is responsible.

# Images Patterns

Choose the workflow that matches the existing storage and delivery architecture, then read its implementation guide.

| Workflow | Documentation |
|----------|---------------|
| Let users upload without exposing account credentials | [Direct Creator Upload](https://developers.cloudflare.com/images/storage/upload-images/direct-creator-upload/) |
| Serve images for different layouts and display densities | [Make responsive images](https://developers.cloudflare.com/images/optimization/make-responsive-images/) |
| Select output format for hosted images | [Hosted-image format optimization](https://developers.cloudflare.com/images/optimization/hosted-images/serve-uploaded-images/#optimize-format) |
| Select output format for a Worker pipeline | [Workers optimization binding](https://developers.cloudflare.com/images/optimization/binding/) |
| Optimize user uploads, add a watermark, and store the result in R2 | [Transform user-uploaded images before uploading to R2](https://developers.cloudflare.com/images/tutorials/optimize-user-uploaded-image/) |
| Compose overlays and watermarks | [Draw overlays](https://developers.cloudflare.com/images/optimization/draw-overlays/) |
| Cache a Worker transformation response | [Binding methods and caching guidance](https://developers.cloudflare.com/images/optimization/binding/#methods) |
| Configure hosted-image browser caching | [Browser TTL](https://developers.cloudflare.com/images/optimization/hosted-images/browser-ttl/) |

Adapt dimensions and quality to the actual layout and representative source images. Keep upload credentials server-side, preserve the application's access checks, and validate both the resulting image and its response headers. Consult [limits and troubleshooting](gotchas.md) before choosing batch sizes or retry behavior.

# Cloudflare Images

Choose the image source and operation before selecting an API. Hosted-image management, remote URL transformations, and the Workers optimization binding have different contracts. Retrieve the documentation for the path the project uses.

| Task | Start here |
|------|------------|
| Optimize image bytes in a Worker or manage hosted images | [API selection](api.md) |
| Configure a binding, variants, or private delivery | [Configuration](configuration.md) |
| Accept client uploads, serve responsive images, watermark, or store results in R2 | [Patterns](patterns.md) |
| Diagnose failures, check limits, or investigate caching | [Troubleshooting](gotchas.md) |

For new work, inspect the project's installed Wrangler version, compatibility settings, existing image storage, and public/private access requirements. Read only the relevant linked pages and adapt them to the project; preserve existing conventions and verify behavior with representative images.

# KV API Reference

Read the relevant API page before implementing; it defines current options, result shapes, supported bulk operations, and constraints.

| Task | Documentation |
|------|---------------|
| Read one or several keys; choose text, JSON, binary, or stream results | [Read key-value pairs](https://developers.cloudflare.com/kv/api/read-key-value-pairs/) |
| Read metadata with values; tune read caching or coalesce related keys | [Read guidance](https://developers.cloudflare.com/kv/api/read-key-value-pairs/) |
| Write values and metadata; set absolute expiration or a relative lifetime | [Write key-value pairs](https://developers.cloudflare.com/kv/api/write-key-value-pairs/) |
| Delete a key | [Delete key-value pairs](https://developers.cloudflare.com/kv/api/delete-key-value-pairs/) |
| Enumerate keys, filter by prefix, and paginate | [List keys](https://developers.cloudflare.com/kv/api/list-keys/) |
| Access namespaces or perform bulk operations outside a Worker | [KV REST API](https://developers.cloudflare.com/api/resources/kv/) and [Wrangler KV commands](https://developers.cloudflare.com/kv/reference/kv-commands/) |

Handle missing values explicitly: JavaScript reads return `null` for absent keys; valid stored values can be falsy. Choose defaults separately from how you handle request failures.

For pagination, follow the returned cursor until `list_complete` is true, even if a page has no keys. Preserve the original prefix on subsequent calls. Listing returns key information, not stored values; use the listing guide to decide whether metadata avoids additional reads.

Use [gotchas.md](./gotchas.md) for consistency and contention decisions before adding retries or read-after-write verification.

# KV Configuration

Read the setup guide for the target environment before creating resources or editing bindings.

| Task | Documentation |
|------|---------------|
| Create a namespace and connect a Worker | [Get started](https://developers.cloudflare.com/kv/get-started/) and [KV bindings](https://developers.cloudflare.com/kv/concepts/kv-bindings/) |
| Configure staging and production namespaces | [KV environments](https://developers.cloudflare.com/kv/reference/environments/) |
| Generate Worker environment and binding types | [Workers TypeScript](https://developers.cloudflare.com/workers/languages/typescript/) |
| Develop against local storage or a remote binding | [KV local development](https://developers.cloudflare.com/kv/concepts/kv-bindings/) and [remote bindings](https://developers.cloudflare.com/workers/local-development/#remote-bindings) |
| Manage namespaces, individual keys, and bulk files from the CLI | [Wrangler KV commands](https://developers.cloudflare.com/kv/reference/kv-commands/) |
| Manage KV from another service or SDK | [KV REST API](https://developers.cloudflare.com/api/resources/kv/) |

Choose the namespace, account, and environment deliberately. Local KV data is separate from remote data; a remote binding accesses the selected Cloudflare namespace even when Worker code runs locally. Check the command's local/remote options and environment selection before seeding or inspecting data. A separate preview namespace is not required simply to use local KV.

Use generated types for binding shapes. JSON type annotations do not validate stored data at runtime; validate application data when its source or schema requires it.

# KV Gotchas & Troubleshooting

Read the linked explanation before applying a workaround.

| Symptom or decision | Documentation and guidance |
|---------------------|----------------------------|
| Stale value after a write or delete | [How KV works](https://developers.cloudflare.com/kv/concepts/how-kv-works/): allow for eventual consistency; neither local read-after-write visibility nor a fixed global propagation deadline is guaranteed. |
| Newly created key still appears absent | [Read caching](https://developers.cloudflare.com/kv/api/read-key-value-pairs/): missing-key lookups are cached too. Treat check-then-create as a race, not an atomic existence test. |
| Concurrent updates overwrite each other or writes are throttled | [Concurrent writes](https://developers.cloudflare.com/kv/api/write-key-value-pairs/#concurrent-writes-to-the-same-key): retries do not make read-modify-write atomic. Use coordination when correctness depends on ordering. |
| Missing-value errors | [Read results](https://developers.cloudflare.com/kv/api/read-key-value-pairs/): distinguish `null` from valid falsy values and distinguish absence from an operation failure. |
| Slow reads, large results, or excessive operations | [Read guidance](https://developers.cloudflare.com/kv/api/read-key-value-pairs/): select result types and bulk reads to match the workload; increasing read cache lifetime trades freshness for cache reuse. |
| Unexpected empty listing page | [Pagination](https://developers.cloudflare.com/kv/api/list-keys/): use the completion flag and cursor, not page length, to determine whether to continue. |
| Data present in one environment but missing in another | [KV bindings](https://developers.cloudflare.com/kv/concepts/kv-bindings/) and [environments](https://developers.cloudflare.com/kv/reference/environments/): check local versus remote storage and the selected namespace. |
| Size, operation, or write-rate failures | [Limits](https://developers.cloudflare.com/kv/platform/limits/): retrieve current constraints before sizing values, batches, or retry policies. |
| Estimate costs or explain billing | [Pricing](https://developers.cloudflare.com/kv/platform/pricing/): check allowances, billable operations, storage, and bulk accounting for the actual workload. |

Confirm freshness and failure requirements before adding a cache or a permissive fallback; see [patterns.md](./patterns.md).

# KV Patterns & Best Practices

Read the guide for the pattern before implementing it, and confirm that [KV's consistency model](https://developers.cloudflare.com/kv/concepts/how-kv-works/) fits the application.

| Task | Documentation and design decision |
|------|-----------------------------------|
| Cache application data or API results | [Cache data with KV](https://developers.cloudflare.com/kv/examples/cache-data-with-workers-kv/): decide acceptable staleness, expiration, and behavior when the origin fails. |
| Cache eligible HTTP responses | [Workers Cache](https://developers.cloudflare.com/workers/cache/): choose the HTTP caching mechanism based on response semantics. |
| Store configuration or feature flags | [Distributed configuration](https://developers.cloudflare.com/kv/examples/distributed-configuration-with-workers-kv/): choose defaults and rollout behavior that tolerate delayed updates. |
| Coalesce related keys | [Read guidance](https://developers.cloudflare.com/kv/api/read-key-value-pairs/): fewer reads can improve cache reuse, but combined values couple updates and can introduce write races. |
| Organize and enumerate keys by prefix | [List keys](https://developers.cloudflare.com/kv/api/list-keys/): use a consistent naming scheme and paginate every listing. |
| Attach schema versions or other metadata | [Write metadata](https://developers.cloudflare.com/kv/api/write-key-value-pairs/) and [read metadata](https://developers.cloudflare.com/kv/api/read-key-value-pairs/): define compatibility and migration behavior for older records; migrations must account for concurrent writes. |

## Application-specific decisions

The linked APIs are building blocks, not complete session or multi-tier cache implementations. Preserve these requirements when designing an application:

- For a memory → KV → origin cache, define each layer's lifetime and refill behavior. Process memory is not shared durable state; KV adds its own stale-value and negative-lookup caching.
- For sessions, decide how quickly creation, updates, and revocation must become visible. KV alone cannot provide immediate global revocation or guaranteed immediate reads after session creation. Use a store with suitable consistency when those are requirements, and define application expiration checks using the [write expiration guidance](https://developers.cloudflare.com/kv/api/write-key-value-pairs/).
- For counters, rate limits, or other atomic read-modify-write decisions, use coordination such as [Durable Objects](https://developers.cloudflare.com/durable-objects/). Serializing writes through an object does not make separate KV reads strongly consistent.
- Choose missing-data defaults separately from service-error handling. A fallback appropriate for display preferences may be inappropriate for authorization or session validation.

# Cloudflare Workers KV

Use KV for read-heavy configuration, preferences, and application caches that tolerate stale data. Read [how KV works](https://developers.cloudflare.com/kv/concepts/how-kv-works/) before choosing it: reads are eventually consistent, including cached missing keys, and immediate visibility is not guaranteed even in the location of a write.

For atomic updates or coordination, consider [Durable Objects](https://developers.cloudflare.com/durable-objects/); for relational queries, [D1](../d1/); for large objects, [R2](../r2/). Use the [storage comparison](https://developers.cloudflare.com/workers/platform/storage-options/) to choose based on requirements.

Read the current documentation for the task before implementing. Use the [KV documentation index](https://developers.cloudflare.com/kv/llms.txt) to discover additional guides; these files preserve task routes rather than copies of APIs, commands, or numeric limits.

## Start here

- [Get started](https://developers.cloudflare.com/kv/get-started/): create a namespace, bind it, and read and write data.
- [configuration.md](./configuration.md): bindings, environments, types, local development, CLI, and REST access.
- [api.md](./api.md): reads, writes, metadata, deletion, bulk operations, and pagination.
- [patterns.md](./patterns.md): caching, sessions, key design, versioning, and fallback decisions.
- [gotchas.md](./gotchas.md): stale reads, concurrent writes, missing values, performance, limits, and pricing.

# Miniflare API

Use the current documentation for method signatures and examples:

| Task | Documentation |
|------|---------------|
| Create, reload, or dispose an instance; wait for its HTTP server | [Get started](https://developers.cloudflare.com/workers/testing/miniflare/get-started/) |
| Dispatch requests and supply request metadata | [Fetch events](https://developers.cloudflare.com/workers/testing/miniflare/core/fetch/) |
| Trigger queue and scheduled handlers programmatically | [Dispatching events](https://developers.cloudflare.com/workers/testing/miniflare/get-started/#dispatching-events) |
| Configure queue producers and consumers | [Queues](https://developers.cloudflare.com/workers/testing/miniflare/core/queues/) |
| Trigger scheduled events over HTTP or the API | [Scheduled events](https://developers.cloudflare.com/workers/testing/miniflare/core/scheduled/) |
| Access bindings from tests | [Interacting with bindings](https://developers.cloudflare.com/workers/testing/miniflare/writing-tests/#interacting-with-bindings) |
| Access local storage | [KV](https://developers.cloudflare.com/workers/testing/miniflare/storage/kv/), [R2](https://developers.cloudflare.com/workers/testing/miniflare/storage/r2/), [D1](https://developers.cloudflare.com/workers/testing/miniflare/storage/d1/), [Durable Objects](https://developers.cloudflare.com/workers/testing/miniflare/storage/durable-objects/), [Cache](https://developers.cloudflare.com/workers/testing/miniflare/storage/cache/) |
| Handle a WebSocket upgrade in a test | [WebSockets](https://developers.cloudflare.com/workers/testing/miniflare/core/web-sockets/) |

For constructor options, read [configuration.md](./configuration.md). For runtime-specific test helpers, read [patterns.md](./patterns.md).

# Miniflare Configuration

Direct Miniflare does not read Wrangler configuration. Configure its bindings explicitly and build TypeScript or bundled Workers before starting tests; see [writing tests](https://developers.cloudflare.com/workers/testing/miniflare/writing-tests/).

Match the Worker's intended compatibility date and flags when testing its behavior. Consult [compatibility dates](https://developers.cloudflare.com/workers/testing/miniflare/core/compatibility/) rather than substituting a fixed date from a sample.

| Configure | Documentation |
|-----------|---------------|
| Script source, HTTP server, request metadata, or reloading | [Get started](https://developers.cloudflare.com/workers/testing/miniflare/get-started/) |
| Module format and resolution rules | [Modules](https://developers.cloudflare.com/workers/testing/miniflare/core/modules/) |
| Values and file-backed bindings | [Variables and secrets](https://developers.cloudflare.com/workers/testing/miniflare/core/variables-secrets/) |
| Service bindings, shared storage, and several Workers | [Multiple Workers](https://developers.cloudflare.com/workers/testing/miniflare/core/multiple-workers/) |
| Storage bindings and documented persistence options | [KV](https://developers.cloudflare.com/workers/testing/miniflare/storage/kv/), [R2](https://developers.cloudflare.com/workers/testing/miniflare/storage/r2/), [D1](https://developers.cloudflare.com/workers/testing/miniflare/storage/d1/), [Durable Objects](https://developers.cloudflare.com/workers/testing/miniflare/storage/durable-objects/), [Cache](https://developers.cloudflare.com/workers/testing/miniflare/storage/cache/) |
| Queue producers and consumers | [Queues](https://developers.cloudflare.com/workers/testing/miniflare/core/queues/) |

If the task is to run tests from the project's build and Wrangler configuration, consider the [integration test harness](https://developers.cloudflare.com/workers/testing/test-harness/) or [Workers Vitest setup](https://developers.cloudflare.com/workers/testing/vitest-integration/write-your-first-test/).

# Miniflare Troubleshooting and Migrations

| Symptom or task | Check |
|-----------------|-------|
| TypeScript, bundled code, or imports fail to load | [Custom builds](https://developers.cloudflare.com/workers/testing/miniflare/writing-tests/#custom-builds) and [module rules](https://developers.cloudflare.com/workers/testing/miniflare/core/modules/#module-rules) |
| Bindings from Wrangler configuration are missing | [Interacting with bindings](https://developers.cloudflare.com/workers/testing/miniflare/writing-tests/#interacting-with-bindings) — direct Miniflare needs explicit configuration |
| Tests disagree with Worker runtime behavior | [Test runtime differences](https://developers.cloudflare.com/workers/testing/miniflare/writing-tests/) and [compatibility dates](https://developers.cloudflare.com/workers/testing/miniflare/core/compatibility/) |
| Instances keep running, ports conflict, or request metadata is unexpected | [Instance lifecycle and HTTP server](https://developers.cloudflare.com/workers/testing/miniflare/get-started/) — dispatching a request without HTTP does not mean the instance has no HTTP server |
| Storage disappears or leaks across tests | Check the relevant [storage configuration](./configuration.md) and the chosen test tool's persistence settings |
| Breakpoints are needed with direct Miniflare | [Attaching a debugger](https://developers.cloudflare.com/workers/testing/miniflare/developing/debugger/) |
| Upgrade a Miniflare 2 application | [Migrate from version 2](https://developers.cloudflare.com/workers/testing/miniflare/migrations/from-v2/) |
| Upgrade an existing Workers Vitest package | [Migrate to Vitest plugin](https://developers.cloudflare.com/workers/testing/vitest-integration/migration-guides/migrate-to-vitest-plugin/) |
| Replace unstable_dev tests | [Migration guide](https://developers.cloudflare.com/workers/testing/vitest-integration/migration-guides/migrate-from-unstable-dev/) and [integration test harness](https://developers.cloudflare.com/workers/testing/test-harness/) |

For a migration, choose the target using [the testing-tool decision](./README.md#choose-the-testing-tool) before translating old options. A historical migration page describes that version transition; use current setup documentation for new test suites.

# Miniflare Testing Patterns

Choose the test runtime before adapting an example. With direct Miniflare, the Worker runs in workerd while the test runner runs in Node.js; importing Worker functions into Node.js can change runtime-dependent behavior. See [writing tests](https://developers.cloudflare.com/workers/testing/miniflare/writing-tests/).

| Task | Documentation |
|------|---------------|
| Write unit tests in the Workers runtime | [Workers Vitest setup](https://developers.cloudflare.com/workers/testing/vitest-integration/write-your-first-test/) |
| Use event, Durable Object, or other runtime test helpers | [Vitest test APIs](https://developers.cloudflare.com/workers/testing/vitest-integration/test-apis/) |
| Test built Workers from an external runner | [Integration test harness](https://developers.cloudflare.com/workers/testing/test-harness/) |
| Build a custom runner with direct simulator control | [Miniflare writing tests](https://developers.cloudflare.com/workers/testing/miniflare/writing-tests/) |
| Access emulated bindings from Node.js | [getPlatformProxy](https://developers.cloudflare.com/workers/wrangler/api/#getplatformproxy) |
| Mock outbound requests in Workers Vitest tests | [Mock outbound requests](https://developers.cloudflare.com/workers/testing/vitest-integration/mock-outbound-requests/) |
| Understand Vitest runtime isolation and concurrency | [Isolation and concurrency](https://developers.cloudflare.com/workers/testing/vitest-integration/isolation-and-concurrency/) |
| Simulate inter-Worker calls and substitute services | [Multiple Workers](https://developers.cloudflare.com/workers/testing/miniflare/core/multiple-workers/) |
| Test WebSockets or access local storage | [API routing](./api.md) |

`getPlatformProxy` is for Node.js callers. The Workers Vitest runtime modules require tests running in the Workers runtime; they are not a substitute for calling `getPlatformProxy` in a Node.js test.

For direct Miniflare, clean up instances after tests using the documented [lifecycle](https://developers.cloudflare.com/workers/testing/miniflare/get-started/#watching-reloading-and-disposing). Choose persistence deliberately so tests do not inherit unintended state.

# Miniflare

Miniflare provides programmatic control of local Workers simulation. Read the linked documentation before choosing APIs, configuration, or a migration path.

## Choose the testing tool

| Need | Start here |
|------|------------|
| Unit tests that execute in the Workers runtime | [Workers Vitest setup](https://developers.cloudflare.com/workers/testing/vitest-integration/write-your-first-test/) |
| Integration tests against built Workers | [Integration test harness](https://developers.cloudflare.com/workers/testing/test-harness/) |
| Low-level simulator control for a custom harness | [Miniflare testing guide](https://developers.cloudflare.com/workers/testing/miniflare/writing-tests/) |
| Binding access from a Node.js process | [Wrangler getPlatformProxy](https://developers.cloudflare.com/workers/wrangler/api/#getplatformproxy) |

For interactive local development, use the project's Wrangler or Cloudflare Vite workflow. Direct Miniflare is useful when the higher-level testing tools do not expose the control needed.

## Read for the task

- [Get started](https://developers.cloudflare.com/workers/testing/miniflare/get-started/) — installation, scripts, lifecycle, and event dispatch.
- [API routing](./api.md) — events and access to local resources.
- [Configuration](./configuration.md) — modules, bindings, compatibility, and multiple Workers.
- [Testing patterns](./patterns.md) — runtime choice, mocking, and test lifecycle.
- [Troubleshooting and migrations](./gotchas.md) — build/configuration differences and existing test suites.

# CNI API Reference

See [README.md](README.md) for overview.

## Base

```
https://api.cloudflare.com/client/v4
Auth: Authorization: Bearer <token>
```

## SDK Namespaces

**Primary (recommended):**
```typescript
client.networkInterconnects.interconnects.*
client.networkInterconnects.cnis.*
client.networkInterconnects.slots.*
```

**Alternate (deprecated):**
```typescript
client.magicTransit.cfInterconnects.*
```

Use `networkInterconnects` namespace for all new code.

## Interconnects

```http
GET    /accounts/{account_id}/cni/interconnects              # Query: page, per_page
POST   /accounts/{account_id}/cni/interconnects              # Query: validate_only=true (optional)
GET    /accounts/{account_id}/cni/interconnects/{icon}
GET    /accounts/{account_id}/cni/interconnects/{icon}/status
GET    /accounts/{account_id}/cni/interconnects/{icon}/loa   # Returns PDF
DELETE /accounts/{account_id}/cni/interconnects/{icon}
```

**Create Body:** `account`, `slot_id`, `type`, `facility`, `speed`, `name`, `description`  
**Status Values:** `active` | `healthy` | `unhealthy` | `pending` | `down`

**Response Example:**
```json
{"result": [{"id": "icon_abc", "name": "prod", "type": "direct", "facility": "EWR1", "speed": "10G", "status": "active"}]}
```

## CNI Objects (BGP config)

```http
GET    /accounts/{account_id}/cni/cnis
POST   /accounts/{account_id}/cni/cnis
GET    /accounts/{account_id}/cni/cnis/{cni}
PUT    /accounts/{account_id}/cni/cnis/{cni}
DELETE /accounts/{account_id}/cni/cnis/{cni}
```

Body: `account`, `cust_ip`, `cf_ip`, `bgp_asn`, `bgp_password`, `vlan`

## Slots

```http
GET /accounts/{account_id}/cni/slots
GET /accounts/{account_id}/cni/slots/{slot}
```

Query: `facility`, `occupied`, `speed`

## Health Checks

Configure via Magic Transit/WAN tunnel endpoints (CNI v2).

```typescript
await client.magicTransit.tunnels.update(accountId, tunnelId, {
  health_check: { enabled: true, target: '192.0.2.1', rate: 'high', type: 'request' },
});
```

Rates: `high` | `medium` | `low`. Types: `request` | `reply`. See [Magic Transit docs](https://developers.cloudflare.com/magic-transit/how-to/configure-tunnel-endpoints/#add-tunnels).

## Settings

```http
GET /accounts/{account_id}/cni/settings
PUT /accounts/{account_id}/cni/settings
```

Body: `default_asn`

## TypeScript SDK

```typescript
import Cloudflare from 'cloudflare';

const client = new Cloudflare({ apiToken: process.env.CF_TOKEN });

// List
await client.networkInterconnects.interconnects.list({ account_id: id });

// Create with validation
await client.networkInterconnects.interconnects.create({
  account_id: id,
  account: id,
  slot_id: 'slot_abc',
  type: 'direct',
  facility: 'EWR1',
  speed: '10G',
  name: 'prod-interconnect',
}, {
  query: { validate_only: true }, // Dry-run validation
});

// Create without validation
await client.networkInterconnects.interconnects.create({
  account_id: id,
  account: id,
  slot_id: 'slot_abc',
  type: 'direct',
  facility: 'EWR1',
  speed: '10G',
  name: 'prod-interconnect',
});

// Status
await client.networkInterconnects.interconnects.get(accountId, iconId);

// LOA (use fetch)
const res = await fetch(`https://api.cloudflare.com/client/v4/accounts/${id}/cni/interconnects/${iconId}/loa`, {
  headers: { Authorization: `Bearer ${token}` },
});
await fs.writeFile('loa.pdf', Buffer.from(await res.arrayBuffer()));

// CNI object
await client.networkInterconnects.cnis.create({
  account_id: id,
  account: id,
  cust_ip: '192.0.2.1/31',
  cf_ip: '192.0.2.0/31',
  bgp_asn: 65000,
  vlan: 100,
});

// Slots (filter by facility and speed)
await client.networkInterconnects.slots.list({
  account_id: id,
  occupied: false,
  facility: 'EWR1',
  speed: '10G',
});
```

## Python SDK

```python
from cloudflare import Cloudflare

client = Cloudflare(api_token=os.environ["CF_TOKEN"])

# List, create, status (same pattern as TypeScript)
client.network_interconnects.interconnects.list(account_id=id)
client.network_interconnects.interconnects.create(account_id=id, account=id, slot_id="slot_abc", type="direct", facility="EWR1", speed="10G")
client.network_interconnects.interconnects.get(account_id=id, icon=icon_id)

# CNI objects and slots
client.network_interconnects.cnis.create(account_id=id, cust_ip="192.0.2.1/31", cf_ip="192.0.2.0/31", bgp_asn=65000)
client.network_interconnects.slots.list(account_id=id, occupied=False)
```

## cURL

```bash
# List interconnects
curl "https://api.cloudflare.com/client/v4/accounts/${ACCOUNT_ID}/cni/interconnects" \
  -H "Authorization: Bearer ${CF_TOKEN}"

# Create interconnect
curl -X POST "https://api.cloudflare.com/client/v4/accounts/${ACCOUNT_ID}/cni/interconnects?validate_only=true" \
  -H "Authorization: Bearer ${CF_TOKEN}" -H "Content-Type: application/json" \
  -d '{"account": "id", "slot_id": "slot_abc", "type": "direct", "facility": "EWR1", "speed": "10G"}'

# LOA PDF
curl "https://api.cloudflare.com/client/v4/accounts/${ACCOUNT_ID}/cni/interconnects/${ICON_ID}/loa" \
  -H "Authorization: Bearer ${CF_TOKEN}" --output loa.pdf
```

## Not Available via API

**Missing Capabilities:**
- BGP session state query (use Dashboard or BGP logs)
- Bandwidth utilization metrics (use external monitoring)
- Traffic statistics per interconnect
- Historical uptime/downtime data
- Light level readings (contact account team)
- Maintenance window scheduling (notifications only)

## Resources

- [API Docs](https://developers.cloudflare.com/api/resources/network_interconnects/)
- [TypeScript SDK](https://github.com/cloudflare/cloudflare-typescript)
- [Python SDK](https://github.com/cloudflare/cloudflare-python)

# CNI Configuration

See [README.md](README.md) for overview.

## Workflow (2-4 weeks)

1. **Submit request** (Week 1): Contact account team, provide type/location/use case
2. **Review config** (Week 1-2, v1 only): Approve IP/VLAN/spec doc
3. **Order connection** (Week 2-3):
   - **Direct**: Get LOA, order cross-connect from facility
   - **Partner**: Order virtual circuit in partner portal
   - **Cloud**: Order Direct Connect/Cloud Interconnect, send LOA+VLAN to CF
4. **Configure** (Week 3): Both sides configure per doc
5. **Test** (Week 3-4): Ping, verify BGP, check routes
6. **Health checks** (Week 4): Configure [Magic Transit](https://developers.cloudflare.com/magic-transit/how-to/configure-tunnel-endpoints/#add-tunnels) or [Magic WAN](https://developers.cloudflare.com/magic-wan/configuration/manually/how-to/configure-tunnel-endpoints/#add-tunnels) health checks
7. **Activate** (Week 4): Route traffic, verify flow
8. **Monitor**: Enable [maintenance notifications](https://developers.cloudflare.com/network-interconnect/monitoring-and-alerts/#enable-cloudflare-status-maintenance-notification)

## BGP Configuration

**v1 Requirements:**
- BGP ASN (provide during setup)
- /31 subnet for peering
- Optional: BGP password

**v2:** Simplified, less BGP config needed.

**BGP over CNI (Dec 2024):** Magic WAN/Transit can now peer BGP directly over CNI v2 (no GRE tunnel required).

**Example v1 BGP:**
```
Router ID: 192.0.2.1
Peer IP: 192.0.2.0
Remote ASN: 13335
Local ASN: 65000
Password: [optional]
VLAN: 100
```

## Cloud Interconnect Setup

### AWS Direct Connect (Beta)

**Requirements:** Magic WAN, AWS Dedicated Direct Connect 1/10 Gbps.

**Process:**
1. Contact CF account team
2. Choose location
3. Order in AWS portal
4. AWS provides LOA + VLAN ID
5. Send to CF account team
6. Wait ~4 weeks

**Post-setup:** Add [static routes](https://developers.cloudflare.com/magic-wan/configuration/manually/how-to/configure-routes/#configure-static-routes) to Magic WAN. Enable [bidirectional health checks](https://developers.cloudflare.com/magic-wan/configuration/manually/how-to/configure-tunnel-endpoints/#legacy-bidirectional-health-checks).

### GCP Cloud Interconnect (Beta)

**Setup via Dashboard:**
1. Interconnects → Create → Cloud Interconnect → Google
2. Provide name, MTU (match GCP VLAN attachment), speed (50M-50G granular options available for partner interconnects)
3. Enter VLAN attachment pairing key
4. Confirm order

**Routing to GCP:** Add [static routes](https://developers.cloudflare.com/magic-wan/configuration/manually/how-to/configure-routes/#configure-static-routes). BGP routes from GCP Cloud Router **ignored**.

**Routing to CF:** Configure [custom learned routes](https://cloud.google.com/network-connectivity/docs/router/how-to/configure-custom-learned-routes) in Cloud Router. Request prefixes from CF account team.

## Monitoring

**Dashboard Status:**

| Status | Meaning |
|--------|---------|
| **Healthy** | Link operational, traffic flowing, health checks passing |
| **Active** | Link up, sufficient light, Ethernet negotiated |
| **Unhealthy** | Link down, no/low light (<-20 dBm), can't negotiate |
| **Pending** | Cross-connect incomplete, device unresponsive, RX/TX swapped |
| **Down** | Physical link down, no connectivity |

**Alerts:**

**CNI Connection Maintenance** (Magic Networking only):
```
Dashboard → Notifications → Add
Product: Cloudflare Network Interconnect
Type: Connection Maintenance Alert
```
Warnings up to 2 weeks advance. 6hr delay for new additions.

**Cloudflare Status Maintenance** (entire PoP):
```
Dashboard → Notifications → Add
Product: Cloudflare Status
Filter PoPs: gru,fra,lhr
```

**Find PoP code:**
```
Dashboard → Magic Transit/WAN → Configuration → Interconnects
Select CNI → Note Data Center (e.g., "gru-b")
Use first 3 letters: "gru"
```

## Best Practices

**Critical config-specific practices:**
- /31 subnets required for BGP
- BGP passwords recommended
- BFD for fast failover (v1 only)
- Test ping connectivity before BGP
- Enable maintenance notifications immediately after activation
- Monitor status programmatically via API

For design patterns, HA architecture, and security best practices, see [patterns.md](./patterns.md).

# CNI Gotchas & Troubleshooting

## Common Errors

### "Status: Pending"

**Cause:** Cross-connect not installed, RX/TX fibers reversed, wrong fiber type, or low light levels
**Solution:**
1. Verify cross-connect installed
2. Check fiber at patch panel
3. Swap RX/TX fibers
4. Check light with optical power meter (target > -20 dBm)
5. Contact account team

### "Status: Unhealthy"

**Cause:** Physical issue, low light (<-20 dBm), optic mismatch, or dirty connectors
**Solution:**
1. Check physical connections
2. Clean fiber connectors
3. Verify optic types (10GBASE-LR/100GBASE-LR4)
4. Test with known-good optics
5. Check patch panel
6. Contact account team

### "BGP Session Down"

**Cause:** Wrong IP addressing, wrong ASN, password mismatch, or firewall blocking TCP/179
**Solution:**
1. Verify IPs match CNI object
2. Confirm ASN correct
3. Check BGP password
4. Verify no firewall on TCP/179
5. Check BGP logs
6. Review BGP timers

### "Low Throughput"

**Cause:** MTU mismatch, fragmentation, single GRE tunnel (v1), or routing inefficiency
**Solution:**
1. Check MTU (1500↓/1476↑ for v1, 1500 both for v2)
2. Test various packet sizes
3. Add more GRE tunnels (v1)
4. Consider upgrading to v2
5. Review routing tables
6. Use LACP for bundling (v1)

## API Errors

### 400 Bad Request: "slot_id already occupied"

**Cause:** Another interconnect already uses this slot  
**Solution:** Use `occupied=false` filter when listing slots:
```typescript
await client.networkInterconnects.slots.list({
  account_id: id,
  occupied: false,
  facility: 'EWR1',
});
```

### 400 Bad Request: "invalid facility code"

**Cause:** Typo or unsupported facility  
**Solution:** Check [locations PDF](https://developers.cloudflare.com/network-interconnect/static/cni-locations-05-may-2026.pdf) for valid codes

### 403 Forbidden: "Enterprise plan required"

**Cause:** Account not enterprise-level  
**Solution:** Contact account team to upgrade

### 422 Unprocessable: "validate_only request failed"

**Cause:** Dry-run validation found issues (wrong slot, invalid config)  
**Solution:** Review error message details, fix config before real creation

### Rate Limiting

**Limit:** 1200 requests/5min per token  
**Solution:** Implement exponential backoff, cache slot listings

## Cloud-Specific Issues

### AWS Direct Connect: "VLAN not matching"

**Cause:** VLAN ID from AWS LOA doesn't match CNI config  
**Solution:**
1. Get VLAN from AWS Console after ordering
2. Send exact VLAN to CF account team
3. Verify match in CNI object config

### AWS: "Connection stuck in Pending"

**Cause:** LOA not provided to CF or AWS connection not accepted  
**Solution:**
1. Verify AWS connection status is "Available"
2. Confirm LOA sent to CF account team
3. Wait for CF team acceptance (can take days)

### GCP: "BGP routes not propagating"

**Cause:** BGP routes from GCP Cloud Router **ignored by design**  
**Solution:** Use [static routes](https://developers.cloudflare.com/magic-wan/configuration/manually/how-to/configure-routes/#configure-static-routes) in Magic WAN instead

### GCP: "Cannot query VLAN attachment status via API"

**Cause:** GCP Cloud Interconnect Dashboard-only (no API yet)  
**Solution:** Check status in CF Dashboard or GCP Console

## Partner Interconnect Issues

### Equinix: "Virtual circuit not appearing"

**Cause:** CF hasn't accepted Equinix connection request  
**Solution:**
1. Verify VC created in Equinix Fabric Portal
2. Contact CF account team to accept
3. Allow 2-3 business days

### Console Connect/Megaport: "API creation fails"

**Cause:** Partner interconnects require partner portal + CF approval  
**Solution:** Cannot fully automate. Order in partner portal, notify CF account team.

## Anti-Patterns

| Anti-Pattern | Why Bad | Solution |
|--------------|---------|----------|
| Single interconnect for production | No SLA, single point of failure | Use ≥2 with device diversity |
| No backup Internet | CNI fails = total outage | Always maintain alternate path |
| Polling status every second | Rate limits, wastes API calls | Poll every 30-60s max |
| Using v1 for Magic WAN v2 workloads | GRE overhead, complexity | Use v2 for simplified routing |
| Assuming BGP session = traffic flowing | BGP up ≠ routes installed | Verify routing tables + test traffic |
| Not enabling maintenance alerts | Surprise downtime during maintenance | Enable notifications immediately |
| Hardcoding VLAN in automation | VLAN assigned by CF (v1) | Get VLAN from CNI object response |
| Using Direct without colocation | Can't access cross-connect | Use Partner or Cloud interconnect |

## What's Not Queryable via API

**Cannot retrieve:**
- BGP session state (use Dashboard or BGP logs)
- Light levels (contact account team)
- Historical metrics (uptime, traffic)
- Bandwidth utilization per interconnect
- Maintenance window schedules (notifications only)
- Fiber path details
- Cross-connect installation status

**Workarounds:**
- External monitoring for BGP state
- Log aggregation for historical data
- Notifications for maintenance windows

## Limits

| Resource/Limit | Value | Notes |
|----------------|-------|-------|
| Max optical distance | 10km | Physical limit |
| MTU (v1) | 1500↓ / 1476↑ | Asymmetric |
| MTU (v2) | 1500 both | Symmetric |
| GRE tunnel throughput | 1 Gbps | Per tunnel (v1) |
| Recovery time | Days | No formal SLA |
| Light level minimum | -20 dBm | Target threshold |
| API rate limit | 1200 req/5min | Per token |
| Health check delay | 6 hours | New maintenance alert subscriptions |

# CNI Patterns

See [README.md](README.md) for overview.

## High Availability

**Critical:** Design for resilience from day one.

**Requirements:**
- Device-level diversity (separate hardware)
- Backup Internet connectivity (no SLA on CNI)
- Network-resilient locations preferred
- Regular failover testing

**Architecture:**
```
Your Network A ──10G CNI v2──> CF CCR Device 1
                                     │
Your Network B ──10G CNI v2──> CF CCR Device 2
                                     │
                            CF Global Network (AS13335)
```

**Capacity Planning:**
- Plan across all links
- Account for failover scenarios
- Your responsibility

## Pattern: Magic Transit + CNI v2

**Use Case:** DDoS protection, private connectivity, no GRE overhead.

```typescript
// 1. Create interconnect
const ic = await client.networkInterconnects.interconnects.create({
  account_id: id,
  type: 'direct',
  facility: 'EWR1',
  speed: '10G',
  name: 'magic-transit-primary',
});

// 2. Poll until active
const status = await pollUntilActive(id, ic.id);

// 3. Configure Magic Transit tunnel via Dashboard/API
```

**Benefits:** 1500 MTU both ways, simplified routing.

## Pattern: Multi-Cloud Hybrid

**Use Case:** AWS/GCP workloads with Cloudflare.

**AWS Direct Connect:**
```typescript
// 1. Order Direct Connect in AWS Console
// 2. Get LOA + VLAN from AWS
// 3. Send to CF account team (no API)
// 4. Configure static routes in Magic WAN

await configureStaticRoutes(id, {
  prefix: '10.0.0.0/8',
  nexthop: 'aws-direct-connect',
});
```

**GCP Cloud Interconnect:**
```
1. Get VLAN attachment pairing key from GCP Console
2. Create via Dashboard: Interconnects → Create → Cloud Interconnect → Google
   - Enter pairing key, name, MTU, speed
3. Configure static routes in Magic WAN (BGP routes from GCP ignored)
4. Configure custom learned routes in GCP Cloud Router
```

**Note:** Dashboard-only. No API/SDK support yet.

## Pattern: Multi-Location HA

**Use Case:** 99.99%+ uptime.

```typescript
// Primary (NY)
const primary = await client.networkInterconnects.interconnects.create({
  account_id: id,
  type: 'direct',
  facility: 'EWR1',
  speed: '10G',
  name: 'primary-ewr1',
});

// Secondary (NY, different hardware)
const secondary = await client.networkInterconnects.interconnects.create({
  account_id: id,
  type: 'direct',
  facility: 'EWR2',
  speed: '10G',
  name: 'secondary-ewr2',
});

// Tertiary (LA, different geography)
const tertiary = await client.networkInterconnects.interconnects.create({
  account_id: id,
  type: 'partner',
  facility: 'LAX1',
  speed: '10G',
  name: 'tertiary-lax1',
});

// BGP local preferences:
// Primary: 200
// Secondary: 150
// Tertiary: 100
// Internet: Last resort
```

## Pattern: Partner Interconnect (Equinix)

**Use Case:** Quick deployment, no colocation.

**Setup:**
1. Order virtual circuit in Equinix Fabric Portal
2. Select Cloudflare as destination
3. Choose facility
4. Send details to CF account team
5. CF accepts in portal
6. Configure BGP

**No API automation** – partner portals managed separately.

## Failover & Security

**Failover Best Practices:**
- Use BGP local preferences for priority
- Configure BFD for fast detection (v1)
- Test regularly with traffic shift
- Document runbooks

**Security:**
- BGP password authentication
- BGP route filtering
- Monitor unexpected routes
- Magic Firewall for DDoS/threats
- Minimum API token permissions
- Rotate credentials periodically

## Decision Matrix

| Requirement | Recommended |
|-------------|-------------|
| Collocated with CF | Direct |
| Not collocated | Partner |
| AWS/GCP workloads | Cloud |
| 1500 MTU both ways | v2 |
| VLAN tagging | v1 |
| Public peering | v1 |
| Simplest config | v2 |
| BFD fast failover | v1 |
| LACP bundling | v1 |

## Resources

- [Magic Transit Docs](https://developers.cloudflare.com/magic-transit/)
- [Magic WAN Docs](https://developers.cloudflare.com/magic-wan/)
- [Argo Smart Routing](https://developers.cloudflare.com/argo-smart-routing/)

# Cloudflare Network Interconnect (CNI)

Private, high-performance connectivity to Cloudflare's network. **Enterprise-only**.

## Connection Types

**Direct**: Physical fiber in shared datacenter. 10/100 Gbps. You order cross-connect.

**Partner**: Virtual via Console Connect, Equinix, Megaport, etc. Managed via partner SDN.

**Cloud**: AWS Direct Connect or GCP Cloud Interconnect. Magic WAN only.

## Dataplane Versions

**v1 (Classic)**: GRE tunnel support, VLAN/BFD/LACP, asymmetric MTU (1500↓/1476↑), peering support.

**v2 (Beta)**: No GRE, 1500 MTU both ways, no VLAN/BFD/LACP yet, ECMP instead.

## Use Cases

- **Magic Transit DSR**: DDoS protection, egress via ISP (v1/v2)
- **Magic Transit + Egress**: DDoS + egress via CF (v1/v2)
- **Magic WAN + Zero Trust**: Private backbone (v1 needs GRE, v2 native)
- **Peering**: Public routes at PoP (v1 only)
- **App Security**: WAF/Cache/LB (v1/v2 over Magic Transit)

## Prerequisites

- Enterprise plan
- IPv4 /24+ or IPv6 /48+ prefixes
- BGP ASN for v1
- See [locations PDF](https://developers.cloudflare.com/network-interconnect/static/cni-locations-05-may-2026.pdf)

## Specs

- /31 point-to-point subnets
- 10km max optical distance
- 10G: 10GBASE-LR single-mode
- 100G: 100GBASE-LR4 single-mode
- **No SLA** (free service)
- Backup Internet required

## Throughput

| Direction | 10G | 100G |
|-----------|-----|------|
| CF → Customer | 10 Gbps | 100 Gbps |
| Customer → CF (peering) | 10 Gbps | 100 Gbps |
| Customer → CF (Magic) | 1 Gbps/tunnel or CNI | 1 Gbps/tunnel or CNI |

## Timeline

2-4 weeks typical. Steps: request → config review → order connection → configure → test → enable health checks → activate → monitor.

## In This Reference
- [configuration.md](./configuration.md) - BGP, routing, setup
- [api.md](./api.md) - API endpoints, SDKs
- [patterns.md](./patterns.md) - HA, hybrid cloud, failover
- [gotchas.md](./gotchas.md) - Troubleshooting, limits

## Reading Order by Task

| Task | Files to Load |
|------|---------------|
| Initial setup | README → configuration.md → api.md |
| Create interconnect via API | api.md → gotchas.md |
| Design HA architecture | patterns.md → README |
| Troubleshoot connection | gotchas.md → configuration.md |
| Cloud integration (AWS/GCP) | configuration.md → patterns.md |
| Monitor + alerts | configuration.md |

## Automation Boundary

**API-Automatable:**
- List/create/delete interconnects (Direct, Partner)
- List available slots
- Get interconnect status
- Download LOA PDF
- Create/update CNI objects (BGP config)
- Query settings

**Requires Account Team:**
- Initial request approval
- AWS Direct Connect setup (send LOA+VLAN to CF)
- GCP Cloud Interconnect final activation
- Partner interconnect acceptance (Equinix, Megaport)
- VLAN assignment (v1)
- Configuration document generation (v1)
- Escalations + troubleshooting support

**Cannot Be Automated:**
- Physical cross-connect installation (Direct)
- Partner portal operations (virtual circuit ordering)
- AWS/GCP portal operations
- Maintenance window coordination

## See Also
- [tunnel](../tunnel/) - Alternative for private network connectivity
- [spectrum](../spectrum/) - Layer 4 proxy for TCP/UDP traffic

# Observability APIs

Fetch the applicable reference for current signatures, field locations, units, authentication, and query syntax. Do not infer the Tail event schema from an OpenTelemetry span or a Logpush record.

| Task | Maintained documentation |
| --- | --- |
| Emit console messages and check supported methods | [Console API](https://developers.cloudflare.com/workers/runtime-apis/console/) |
| Filter, group, and aggregate stored Workers Logs | [Query Builder](https://developers.cloudflare.com/workers/observability/query-builder/) |
| Query built-in Workers metrics with GraphQL | [Querying Workers metrics](https://developers.cloudflare.com/analytics/graphql-api/tutorials/querying-workers-metrics/) |
| Define Analytics Engine fields and call `writeDataPoint()` | [Write data points](https://developers.cloudflare.com/analytics/analytics-engine/get-started/) |
| Authenticate and query Analytics Engine datasets | [SQL API](https://developers.cloudflare.com/analytics/analytics-engine/sql-api/) |
| Calculate counts, sums, and averages on sampled events | [Analytics Engine sampling](https://developers.cloudflare.com/analytics/analytics-engine/sampling/) |
| Implement a Tail consumer and inspect event properties | [Tail handler API](https://developers.cloudflare.com/workers/runtime-apis/handlers/tail/) |
| Create and manage Logpush jobs | [Logpush API configuration](https://developers.cloudflare.com/logs/logpush/logpush-job/api-configuration/) |
| Select exported Workers event fields | [Workers Trace Events dataset](https://developers.cloudflare.com/logs/logpush/logpush-job/datasets/account/workers_trace_events/) |
| Export OTLP logs and traces | [OpenTelemetry export](https://developers.cloudflare.com/workers/observability/exporting-opentelemetry-data/) |

Keep dataset field meanings and units consistent between writers and queries. Follow the SQL reference linked from the SQL API for supported date bucketing and aggregate functions; do not assume another SQL dialect's syntax works here. Account for sampling in averages as well as counts and sums.

See [configuration.md](configuration.md) for setup and [patterns.md](patterns.md) for application-level decisions.

# Observability Configuration

Fetch the relevant guide before configuring the selected Worker and deployment environment.

| Task | Maintained documentation |
| --- | --- |
| Enable persisted logs, structured JSON logging, and sampling | [Workers Logs](https://developers.cloudflare.com/workers/observability/logs/workers-logs/) |
| Enable traces and set their sampling independently of logs | [Workers Traces](https://developers.cloudflare.com/workers/observability/traces/) |
| Configure a named deployment environment | [Wrangler environments](https://developers.cloudflare.com/workers/wrangler/environments/) and the environment example in [Workers Logs](https://developers.cloudflare.com/workers/observability/logs/workers-logs/) |
| Bind an Analytics Engine dataset and write its first data point | [Analytics Engine get started](https://developers.cloudflare.com/analytics/analytics-engine/get-started/) |
| Connect a producer to a Tail Worker | [Configure Tail Workers](https://developers.cloudflare.com/workers/observability/logs/tail-workers/) |
| Create a Logpush job, configure access, and enable Worker log delivery | [Workers Logpush](https://developers.cloudflare.com/workers/observability/logs/logpush/) |
| Configure OTLP destinations, authentication, and local persistence | [Exporting OpenTelemetry data](https://developers.cloudflare.com/workers/observability/exporting-opentelemetry-data/) |

## Setup decisions

- Confirm which account, Worker, and environment will emit telemetry, then deploy that configuration and generate representative traffic.
- Decide log and trace sampling separately. Increasing sampling during an investigation changes volume and cost; restore the intended operational settings afterwards.
- For Tail Workers, configure the consumer relationship on the producer Worker; use the guide for deployment order and the handler contract.
- Choose whether to persist data in Cloudflare as well as exporting it. Verify destination names, supported signal types, and credentials using the export guide.
- Use stable structured fields and redact secrets and unnecessary personal data before emission. Configure development and production collection intentionally.

See [gotchas.md](gotchas.md) when configured telemetry is missing.

# Observability Troubleshooting and Constraints

## Missing or incomplete data

| Symptom | Check and authoritative guide |
| --- | --- |
| Logs missing from the dashboard | Confirm the deployed Worker/environment, collection and persistence settings, recent traffic, query time range, and sampling. Follow [Workers Logs](https://developers.cloudflare.com/workers/observability/logs/workers-logs/) and [Query Builder](https://developers.cloudflare.com/workers/observability/query-builder/). |
| Live logs differ from stored logs | Confirm which workflow is being inspected; live streams can sample under load. See [real-time logs](https://developers.cloudflare.com/workers/observability/logs/real-time-logs/). |
| Traces missing or incomplete | Check trace enablement and sampling separately from logs, then consult [tracing setup](https://developers.cloudflare.com/workers/observability/traces/) and [known limitations](https://developers.cloudflare.com/workers/observability/traces/known-limitations/). |
| Export destination has no data | Check signal type, destination name, credentials, endpoint compatibility, and provider status using [OpenTelemetry export](https://developers.cloudflare.com/workers/observability/exporting-opentelemetry-data/). For a Logpush job, use [Workers Logpush](https://developers.cloudflare.com/workers/observability/logs/logpush/). |
| Tail consumer receives no events | Check the producer's consumer configuration and deployment using [Tail Workers](https://developers.cloudflare.com/workers/observability/logs/tail-workers/). |
| Analytics Engine totals or averages look wrong | Account for sample weights and consistent field meanings using [sampling guidance](https://developers.cloudflare.com/analytics/analytics-engine/sampling/). Check [limits](https://developers.cloudflare.com/analytics/analytics-engine/limits/) for missing writes or expired data. |
| Very short operations appear to take no time | Read [performance and timers](https://developers.cloudflare.com/workers/runtime-apis/performance/) and [trace limitations](https://developers.cloudflare.com/workers/observability/traces/known-limitations/). Tracing does not eliminate the runtime's timing restrictions. |

## Limits, retention, and cost

Fetch these pages when estimating cost or diagnosing truncation and missing data:

- [Workers Logs](https://developers.cloudflare.com/workers/observability/logs/workers-logs/): log size, retention, sampling, and pricing.
- [Workers Traces](https://developers.cloudflare.com/workers/observability/traces/) and [known limitations](https://developers.cloudflare.com/workers/observability/traces/known-limitations/): availability, propagation, and instrumentation constraints.
- [Workers pricing](https://developers.cloudflare.com/workers/platform/pricing/): current observability and Tail Worker billing terms.
- [Analytics Engine limits](https://developers.cloudflare.com/analytics/analytics-engine/limits/) and [pricing](https://developers.cloudflare.com/analytics/analytics-engine/pricing/): field/write limits, retention, query costs, and billing availability.
- [Workers Logpush](https://developers.cloudflare.com/workers/observability/logs/logpush/): Workers-specific eligibility, permissions, and pricing.

Sampling reduces coverage as well as volume. Do not interpret the absence of a sampled event as proof that an error did not happen. Keep required diagnostic context while avoiding credentials, full sensitive URLs, and unnecessary personal data in logs, custom dimensions, and exported records.

# Observability Patterns

Use these decisions alongside the linked implementation guides. Application event schemas, billing policies, alert thresholds, and delivery behavior still need to be designed for the application.

| Task | Design decision and documentation |
| --- | --- |
| Usage-based billing | Define the billable event, tenant identity, time window, and accuracy requirements. Start with the maintained [Analytics Engine billing recipe](https://developers.cloudflare.com/analytics/analytics-engine/recipes/usage-based-billing-for-your-saas-product/) and [sampling guidance](https://developers.cloudflare.com/analytics/analytics-engine/sampling/); assess whether sampled estimates satisfy the billing contract. |
| Performance monitoring | Use [built-in metrics](https://developers.cloudflare.com/workers/observability/metrics-and-analytics/) for aggregate health and [traces](https://developers.cloudflare.com/workers/observability/traces/) to investigate dependency latency. Custom measurements need consistent units and aggregation semantics; check [runtime timers](https://developers.cloudflare.com/workers/runtime-apis/performance/) before measuring CPU-only work. |
| Error tracking | Emit structured context without secrets, investigate with [Query Builder](https://developers.cloudflare.com/workers/observability/query-builder/), and choose [Tail Workers](https://developers.cloudflare.com/workers/observability/logs/tail-workers/) if custom alert processing is needed. Define thresholds and duplicate handling for your alert destination. |
| Multi-tenant tracking | Choose the tenant dimension and consistent field positions using [Analytics Engine get started](https://developers.cloudflare.com/analytics/analytics-engine/get-started/) and [sampling guidance](https://developers.cloudflare.com/analytics/analytics-engine/sampling/). Enforce tenant authorization in the application that exposes analytics; a dataset index is not an access-control boundary. |
| Tail Worker filtering | Use the current [Tail handler schema](https://developers.cloudflare.com/workers/runtime-apis/handlers/tail/) for outcomes, exceptions, and timing fields, with [Tail configuration](https://developers.cloudflare.com/workers/observability/logs/tail-workers/) for producer wiring. Define filtering, redaction, and downstream failure handling for the destination. |
| OpenTelemetry export | Prefer the maintained [OTLP export integration](https://developers.cloudflare.com/workers/observability/exporting-opentelemetry-data/) for supported destinations. For Honeycomb, follow [Export to Honeycomb](https://developers.cloudflare.com/workers/observability/exporting-opentelemetry-data/honeycomb/) instead of synthesizing spans from Tail events. |

Use [Workers Logpush](https://developers.cloudflare.com/workers/observability/logs/logpush/) when the requirement is Workers Trace Event delivery to a supported log destination. Use a Tail Worker when custom processing is required beyond the configured export integration.

# Cloudflare Observability

Use this reference to choose a telemetry signal and find the maintained implementation guide. Fetch the linked documentation before writing configuration, queries, or export code; it is the source of truth for APIs, availability, retention, limits, and pricing.

## Choose a signal

| Need | Start here |
| --- | --- |
| Store, search, and investigate historical Worker logs | [Workers Logs](https://developers.cloudflare.com/workers/observability/logs/workers-logs/) |
| Watch a deployment or reproduce an issue live | [Real-time logs and Wrangler tail](https://developers.cloudflare.com/workers/observability/logs/real-time-logs/) |
| Understand request flows and dependency latency | [Workers Traces](https://developers.cloudflare.com/workers/observability/traces/) |
| Monitor built-in request, error, and CPU metrics | [Metrics and analytics](https://developers.cloudflare.com/workers/observability/metrics-and-analytics/) |
| Record custom events and tenant-level usage for SQL analysis | [Analytics Engine](https://developers.cloudflare.com/analytics/analytics-engine/get-started/) |
| Export logs and traces to an observability provider | [OpenTelemetry export](https://developers.cloudflare.com/workers/observability/exporting-opentelemetry-data/) |
| Apply custom filtering, transformation, or delivery logic | [Tail Workers](https://developers.cloudflare.com/workers/observability/logs/tail-workers/) |
| Deliver Workers Trace Events to a supported log storage destination | [Workers Logpush](https://developers.cloudflare.com/workers/observability/logs/logpush/) |

Workers Logs supports retained historical data; live tailing is a separate debugging workflow. Choose persistence, sampling, and export destinations deliberately rather than assuming that every signal is stored or included without usage charges.

## Load only what the task needs

- [configuration.md](configuration.md): enable collection, bindings, environments, and exports.
- [api.md](api.md): logging, telemetry types, SQL, GraphQL, and Logpush APIs.
- [patterns.md](patterns.md): billing, performance, errors, tenant tracking, and export decisions.
- [gotchas.md](gotchas.md): missing data, sampling, timing, privacy, and cost checks.

For broader product tasks, see [Analytics Engine](../analytics-engine/README.md), [GraphQL API](../graphql-api/README.md), and [Tail Workers](../tail-workers/README.md).

# Pages Request Handling

Use the current Pages API and the project framework guide rather than translating a generic Worker example into Pages.

| Task | Documentation |
| --- | --- |
| Implement handlers and use context or asset fallback | [Functions API reference](https://developers.cloudflare.com/pages/functions/api-reference/) |
| Resolve dynamic paths and invocation routes | [Routing](https://developers.cloudflare.com/pages/functions/routing/) |
| Apply shared request logic | [Middleware](https://developers.cloudflare.com/pages/functions/middleware/) |
| Understand framework-generated Worker output | [Advanced mode](https://developers.cloudflare.com/pages/functions/advanced-mode/) |
| Find the guide for the existing framework | [Framework guides](https://developers.cloudflare.com/pages/framework-guides/) |

First identify whether routing comes from the Functions directory or generated advanced-mode output. See [Pages Functions APIs](../pages-functions/api.md) for focused handler tasks and [patterns.md](./patterns.md) for migration decisions.

# Pages Configuration

Inspect the existing project configuration and build output before changing deployment settings.

| Task | Documentation |
| --- | --- |
| Set build commands, root/output directories, and build variables | [Build configuration](https://developers.cloudflare.com/pages/configuration/build-configuration/) |
| Manage Wrangler configuration, environments, and dashboard migration | [Functions configuration](https://developers.cloudflare.com/pages/functions/wrangler-configuration/) |
| Configure resource bindings, variables, and secrets | [Bindings](https://developers.cloudflare.com/pages/functions/bindings/) |
| Set static response headers | [Headers](https://developers.cloudflare.com/pages/configuration/headers/) |
| Configure static redirects and rewrites | [Redirects](https://developers.cloudflare.com/pages/configuration/redirects/) |
| Choose which requests invoke Functions | [Routing](https://developers.cloudflare.com/pages/functions/routing/) |
| Configure monorepo project boundaries | [Monorepos](https://developers.cloudflare.com/pages/configuration/monorepos/) |
| Run the project locally | [Local development](https://developers.cloudflare.com/pages/functions/local-development/) |

Distinguish build-time variables from runtime bindings, and check preview and production separately. Determine whether the framework owns generated routing files before editing them. See [Pages Functions configuration](../pages-functions/configuration.md) for types and placement.

# Pages Troubleshooting

Identify whether the failure occurs during the build, asset serving, or Function execution before changing configuration.

| Task | Documentation |
| --- | --- |
| Build output is missing or incorrect | [Build configuration](https://developers.cloudflare.com/pages/configuration/build-configuration/) |
| A static URL redirects or returns an unexpected 404 | [Serving Pages](https://developers.cloudflare.com/pages/configuration/serving-pages/) |
| Static response headers are not applied | [Headers](https://developers.cloudflare.com/pages/configuration/headers/) |
| A redirect rule does not match | [Redirects](https://developers.cloudflare.com/pages/configuration/redirects/) |
| Investigate a failed Function request | [Debugging and logging](https://developers.cloudflare.com/pages/functions/debugging-and-logging/) |
| Check deployment and file capacity | [Pages limits](https://developers.cloudflare.com/pages/platform/limits/) |
| Understand Function versus static request billing | [Functions pricing](https://developers.cloudflare.com/pages/functions/pricing/) |

Compare the same route in local, preview, and production environments; record the build output and configuration used by each. See [Pages Functions troubleshooting](../pages-functions/gotchas.md) for handler and binding issues.

## Framework-Specific

Fetch the relevant [framework guide](https://developers.cloudflare.com/pages/framework-guides/) before changing adapters or recommending another host. For a move to Workers, follow the [migration guide](https://developers.cloudflare.com/workers/static-assets/migration-guides/migrate-from-pages/).

# Pages Project Decisions

Keep project-specific choices here; fetch framework adapters, configuration, and code examples from the docs.

| Task | Documentation |
| --- | --- |
| Maintain a framework deployment | [Framework guides](https://developers.cloudflare.com/pages/framework-guides/) |
| Deploy from an external build pipeline | [Direct Upload](https://developers.cloudflare.com/pages/get-started/direct-upload/) |
| Manage multiple apps in one repository | [Monorepos](https://developers.cloudflare.com/pages/configuration/monorepos/) |
| Evaluate backend locality | [Smart Placement](https://developers.cloudflare.com/pages/functions/smart-placement/) |
| Plan a move to Workers | [Migrate from Pages to Workers](https://developers.cloudflare.com/workers/static-assets/migration-guides/migrate-from-pages/) |

Check the existing build command, adapter, and output ownership together. Evaluate placement using the application’s backend dependencies and measured latency. For a migration, inventory routes, middleware, bindings, static rules, and deployment settings before following the migration guide.

See [Pages Functions patterns](../pages-functions/patterns.md) for request-level decisions.

# Cloudflare Pages

Use this reference when maintaining an existing Pages project. For new applications, start with Workers as recommended in the [Pages framework guidance](https://developers.cloudflare.com/pages/framework-guides/). Fetch current documentation before implementing.

| Task | Documentation |
| --- | --- |
| Configure the existing build | [Build configuration](https://developers.cloudflare.com/pages/configuration/build-configuration/) |
| Manage automatic deployments from a repository | [Git integration](https://developers.cloudflare.com/pages/configuration/git-integration/) |
| Deploy prebuilt output | [Direct Upload](https://developers.cloudflare.com/pages/get-started/direct-upload/) |
| Implement server-side requests | [Functions API reference](https://developers.cloudflare.com/pages/functions/api-reference/) |
| Plan a move to Workers | [Migrate from Pages to Workers](https://developers.cloudflare.com/workers/static-assets/migration-guides/migrate-from-pages/) |

## In This Reference

- [configuration.md](./configuration.md) — build output, environments, and static rules
- [api.md](./api.md) — request handling and framework integration
- [patterns.md](./patterns.md) — project decisions and migration
- [gotchas.md](./gotchas.md) — build, routing, and deployment investigation

See [Pages Functions](../pages-functions/README.md) for handler-focused navigation. Identify the existing deployment method and framework before proposing changes.

# Pages Functions APIs

Fetch the API reference before writing a handler; keep runtime types and examples in their authoritative documentation.

| Task | Documentation |
| --- | --- |
| Choose method handlers and access EventContext | [API reference](https://developers.cloudflare.com/pages/functions/api-reference/) |
| Read parameters and resolve dynamic routes | [Routing](https://developers.cloudflare.com/pages/functions/routing/) |
| Compose middleware and continue a request | [Middleware](https://developers.cloudflare.com/pages/functions/middleware/) |
| Use a supported resource binding | [Bindings](https://developers.cloudflare.com/pages/functions/bindings/) |
| Handle requests through generated Worker output | [Advanced mode](https://developers.cloudflare.com/pages/functions/advanced-mode/) |

Decide which handler owns the response, where shared state is established, and which paths should fall through to assets. The API reference also covers asynchronous work and asset fetching. See [configuration.md](./configuration.md) for binding setup and [patterns.md](./patterns.md) for request design.

# Pages Functions Configuration

Read Pages-specific configuration before reusing settings from a Worker project.

| Task | Documentation |
| --- | --- |
| Manage Wrangler settings and environment overrides | [Functions configuration](https://developers.cloudflare.com/pages/functions/wrangler-configuration/) |
| Configure supported bindings, variables, and secrets | [Bindings](https://developers.cloudflare.com/pages/functions/bindings/) |
| Generate and configure runtime and environment types | [TypeScript](https://developers.cloudflare.com/pages/functions/typescript/) |
| Run assets and Functions locally | [Local development](https://developers.cloudflare.com/pages/functions/local-development/) |
| Set Function invocation routes | [Routing](https://developers.cloudflare.com/pages/functions/routing/) |
| Evaluate and enable placement | [Smart Placement](https://developers.cloudflare.com/pages/functions/smart-placement/) |

Identify the target deployment environment and which configuration source controls it. Verify Pages support for each binding and the documented local-development behavior before accessing remote resources. See [Pages configuration](../pages/configuration.md) for build output, headers, and redirects.

# Pages Functions Troubleshooting

Start with the request path, deployment environment, and generated output that actually handled the request.

| Task | Documentation |
| --- | --- |
| A Function does not run or receives unexpected parameters | [Routing](https://developers.cloudflare.com/pages/functions/routing/) |
| Middleware is skipped or static fallback fails | [Advanced mode](https://developers.cloudflare.com/pages/functions/advanced-mode/) |
| Middleware order or scope is incorrect | [Middleware](https://developers.cloudflare.com/pages/functions/middleware/) |
| Bindings or secrets differ between environments | [Bindings](https://developers.cloudflare.com/pages/functions/bindings/) |
| Runtime or environment types do not match | [TypeScript](https://developers.cloudflare.com/pages/functions/typescript/) |
| A local request behaves differently | [Local development](https://developers.cloudflare.com/pages/functions/local-development/) |
| Inspect exceptions and deployment logs | [Debugging and logging](https://developers.cloudflare.com/pages/functions/debugging-and-logging/) |
| Check runtime quotas | [Workers limits](https://developers.cloudflare.com/workers/platform/limits/) |
| Check request costs | [Functions pricing](https://developers.cloudflare.com/pages/functions/pricing/) |

Reproduce a failing path through the actual application rather than only calling a handler with a hand-built context. Check the deployed configuration and generated output before changing application code. See [Pages troubleshooting](../pages/gotchas.md) for build, asset, and framework issues.

# Pages Functions Request Design

Choose where behavior belongs before adapting an example.

| Task | Documentation |
| --- | --- |
| Share authentication, logging, and error handling | [Middleware](https://developers.cloudflare.com/pages/functions/middleware/) |
| Choose request handlers and asynchronous completion behavior | [API reference](https://developers.cloudflare.com/pages/functions/api-reference/) |
| Keep asset requests outside Function invocation where appropriate | [Routing](https://developers.cloudflare.com/pages/functions/routing/) |
| Use custom or framework-generated routing | [Advanced mode](https://developers.cloudflare.com/pages/functions/advanced-mode/) |
| Integrate storage or another service | [Bindings](https://developers.cloudflare.com/pages/functions/bindings/) |
| Exercise the assembled application locally | [Local development](https://developers.cloudflare.com/pages/functions/local-development/) |

Define the response owner and middleware scope before adding authentication or response transformations. Test protected routes, rejected requests, and static fallbacks together. Choose consistency and concurrency requirements before using storage for session state or rate limiting; a generic read-modify-write example is not a complete policy.

See [Pages project decisions](../pages/patterns.md) for framework and migration work.

# Cloudflare Pages Functions

Use this reference for server-side behavior in an existing Pages project. For new applications, follow the Workers recommendation in the [Pages framework guidance](https://developers.cloudflare.com/pages/framework-guides/).

| Task | Documentation |
| --- | --- |
| Identify filesystem routes and invocation boundaries | [Routing](https://developers.cloudflare.com/pages/functions/routing/) |
| Implement request handlers | [API reference](https://developers.cloudflare.com/pages/functions/api-reference/) |
| Understand generated Worker output | [Advanced mode](https://developers.cloudflare.com/pages/functions/advanced-mode/) |

Inspect whether the project uses a Functions directory or framework-generated advanced mode before selecting a routing approach. Fetch current documentation for signatures, supported bindings, configuration, and examples.

## In This Reference

- [api.md](./api.md) — handlers, context, middleware, and assets
- [configuration.md](./configuration.md) — bindings, environments, types, and local development
- [patterns.md](./patterns.md) — request ownership and shared logic
- [gotchas.md](./gotchas.md) — route, binding, and runtime investigation

See [Pages](../pages/README.md) for builds and deployment decisions.

# Pipelines API Reference

Code templates and verified behavior. For the full SQL function set and HTTP status semantics, pull `https://developers.cloudflare.com/pipelines/sql-reference/` and the streams docs.

## Worker Binding Interface

```typescript
// from cloudflare:pipelines / @cloudflare/workers-types
interface Pipeline<T = any> { send(records: T[]): Promise<void>; }

interface Env { MY_STREAM: Pipeline; }

export default {
  async fetch(req: Request, env: Env, ctx: ExecutionContext): Promise<Response> {
    await env.MY_STREAM.send([{ event_id: crypto.randomUUID(), amount: 29.99 }]);
    return new Response("OK");
  }
} satisfies ExportedHandler<Env>;
```

- `send()` takes an **array**, returns `Promise<void>` (no confirmation payload).
- Throws on network errors — wrap in try/catch or use `ctx.waitUntil()` for fire-and-forget.
- Validation errors are **not** thrown here (deferred during processing — see [gotchas.md](gotchas.md)).
- Payload/rate limits apply — check `https://developers.cloudflare.com/pipelines/platform/limits/` before sizing batches.

## HTTP Ingest

```
https://{stream-id}.ingest.cloudflare.com
```

Get `{stream-id}` from `npx wrangler pipelines streams list`.

```bash
# Batch (preferred)
curl -X POST https://{stream-id}.ingest.cloudflare.com \
  -H "Content-Type: application/json" \
  -d '[{"event_id":"evt-1","amount":29.99},{"event_id":"evt-2","amount":14.99}]'

# Single event — auto-wrapped in an array
curl -X POST https://{stream-id}.ingest.cloudflare.com \
  -H "Content-Type: application/json" -d '{"event_id":"evt-3","amount":9.99}'
```

If stream auth is enabled, add `-H "Authorization: Bearer $TOKEN"` (token needs **Workers Pipelines Send**). Standard HTTP status codes apply (400 invalid, 401 auth, 413 too large, 429 rate-limited, 5xx retry).

> **JSON only** — no Avro, Protobuf, or CSV input.

## REST Management API

Base: `https://api.cloudflare.com/client/v4/accounts/$ACCOUNT_ID/pipelines/v1`

```bash
# List
curl -s "$BASE_URL/streams"   -H "Authorization: Bearer $API_TOKEN"
curl -s "$BASE_URL/sinks"     -H "Authorization: Bearer $API_TOKEN"
curl -s "$BASE_URL/pipelines" -H "Authorization: Bearer $API_TOKEN"

# Get one (pipeline GET includes status + failure_reason — useful for debugging)
curl -s "$BASE_URL/pipelines/{pipeline-id}" -H "Authorization: Bearer $API_TOKEN"

# Delete in reverse order: pipeline → sink → stream
curl -X DELETE "$BASE_URL/pipelines/{id}" -H "Authorization: Bearer $API_TOKEN"
curl -X DELETE "$BASE_URL/sinks/{id}"     -H "Authorization: Bearer $API_TOKEN"
curl -X DELETE "$BASE_URL/streams/{id}"   -H "Authorization: Bearer $API_TOKEN"
```

> `wrangler pipelines delete` defaults to "no" non-interactively — use the REST API for automated cleanup. Deleting a stream removes buffered events and dependent pipelines.

### Pipeline Lifecycle States

| Status | Meaning |
|--------|---------|
| `running` | Active, processing events |
| `initializing` | Starting up (minutes after creation or recovery) |
| `failed` | Stopped on error — check `failure_reason` (expired token, deleted bucket, disabled catalog) |

> A `GET` on a sink shows `schema.fields: []` — expected. The sink inherits schema from the stream via the pipeline SQL.

## Pipeline SQL (Transforms)

Row-level only — no GROUP BY/aggregation. CTEs (`WITH`) and `UNNEST` are supported. Full function list: `https://developers.cloudflare.com/pipelines/sql-reference/`.

```sql
-- Passthrough / filter / enrich
INSERT INTO my_sink SELECT * FROM my_stream;
INSERT INTO my_sink SELECT * FROM my_stream WHERE amount > 10;
INSERT INTO my_sink
SELECT event_id, UPPER(category) AS category, amount * 1.1 AS amount_with_tax
FROM my_stream;

-- CTE
WITH filtered AS (SELECT event_id, amount FROM my_stream WHERE amount > 50)
INSERT INTO my_sink SELECT * FROM filtered;

-- UNNEST arrays (one per SELECT)
SELECT UNNEST(tags) AS tag FROM my_stream;
```

Supported categories: string, regex, hashing (`sha256`), JSON extraction, timestamp conversion, conditional (`CASE`), `CAST`, `COALESCE`, math/comparison operators.

## Verifying End-to-End Data Flow

```bash
# 1. Pipeline running (not initializing/failed)?
curl -s "$BASE_URL/pipelines/{id}" -H "Authorization: Bearer $API_TOKEN"

# 2. Table created yet? (3–7 min on first flush)
curl -s "https://api.cloudflare.com/client/v4/accounts/$ACCOUNT_ID/r2-catalog/$BUCKET/namespaces/my_ns/tables" \
  -H "Authorization: Bearer $API_TOKEN"

# 3. Data present? (R2 SQL)
curl -s -X POST \
  "https://api.sql.cloudflarestorage.com/api/v1/accounts/$ACCOUNT_ID/r2-sql/query/$BUCKET" \
  -H "Authorization: Bearer $API_TOKEN" -H "Content-Type: application/json" \
  -d '{"query": "SELECT COUNT(*) AS total FROM my_ns.my_table"}'
```

> Expect **3–7 minutes** from first send to first queryable data. Subsequent flushes are much faster.

## See Also

- [configuration.md](configuration.md) — creating resources · [patterns.md](patterns.md) — producers, Logpush, observability
- [r2-sql/api.md](../r2-sql/api.md) — querying results

# Pipelines Configuration

Templates for creating streams, sinks, and pipelines via CLI, REST, or Terraform. For the full flag/field list and allowed values, pull `https://developers.cloudflare.com/pipelines/reference/wrangler-commands/` and the streams/sinks/pipelines docs.

## Naming Rules

- **Streams, sinks, pipelines** use underscores: `my_stream`, `my_sink`, `my_pipeline`.
- **Buckets** use hyphens: `my-bucket`.

## Schema (Structured Streams)

Schema is a JSON object with a `fields` array; each field has `name`, `type`, `required`.

```json
{
  "fields": [
    { "name": "event_id", "type": "string", "required": true },
    { "name": "amount", "type": "float64", "required": false }
  ]
}
```

Field types include `string`, `bool`, `int32/64`, `float32/64`, `timestamp`, `json`, `binary`, `list`, `struct` (with nested `items`/`fields`). For the authoritative type list, see `https://developers.cloudflare.com/pipelines/sql-reference/sql-data-types/`.

Unstructured streams (no schema) store everything in a single `value` column.

> Pipelines auto-adds `__ingest_ts` (TIMESTAMP, day-partitioned). Do **not** include it in your schema.

## Option A: Interactive (Simplest)

```bash
npx wrangler pipelines setup   # creates stream + sink + pipeline, optionally bucket + catalog
```

## Option B: Wrangler CLI (Explicit)

```bash
# 1. Stream
npx wrangler pipelines streams create my_stream --schema-file schema.json

# 2. Sink — R2 Data Catalog (Iceberg). Creates the namespace + table.
npx wrangler pipelines sinks create my_sink \
  --type r2-data-catalog \
  --bucket my-bucket --namespace my_namespace --table my_table \
  --catalog-token $API_TOKEN \
  --compression zstd --roll-interval 300

# 2b. Sink — R2 raw Parquet (alternative)
npx wrangler pipelines sinks create my_sink \
  --type r2 --bucket my-bucket --format parquet \
  --path analytics/events --partitioning "year=%Y/month=%m/day=%d" \
  --access-key-id $KEY --secret-access-key $SECRET

# 3. Pipeline (SQL connects stream → sink)
npx wrangler pipelines create my_pipeline \
  --sql "INSERT INTO my_sink SELECT * FROM my_stream"
```

Tuning knobs (`--compression`, `--roll-interval`, `--roll-size`, etc.) and their allowed values/defaults change — pull the wrangler-commands and sinks docs rather than hardcoding. Rule of thumb: prod `--roll-interval 300+`, dev `10` (creates many small files).

> **⚠️ Pipelines are immutable.** SQL, schema, and sink config can't be changed — delete and recreate.

## Option C: REST API (Programmatic)

Base: `https://api.cloudflare.com/client/v4/accounts/$ACCOUNT_ID/pipelines/v1`

```bash
# Stream
curl -X POST "$BASE_URL/streams" -H "Authorization: Bearer $API_TOKEN" \
  -H "Content-Type: application/json" -d '{
    "name": "my_stream",
    "http": {"enabled": true, "authentication": false},
    "schema": {"fields": [{"name": "event_id", "type": "string", "required": true}]}
  }'

# Sink — NOTE REST field names differ from CLI flags (see table)
curl -X POST "$BASE_URL/sinks" -H "Authorization: Bearer $API_TOKEN" \
  -H "Content-Type: application/json" -d '{
    "name": "my_sink", "type": "r2_data_catalog",
    "config": {"bucket": "my-bucket", "namespace": "my_namespace",
               "table_name": "my_table", "token": "'$API_TOKEN'",
               "rolling_policy": {"interval_seconds": 300}},
    "format": {"type": "parquet"}
  }'

# Pipeline
curl -X POST "$BASE_URL/pipelines" -H "Authorization: Bearer $API_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"name": "my_pipeline", "sql": "INSERT INTO my_sink SELECT * FROM my_stream;"}'
```

**REST field names ≠ CLI flags** (common failure — not obvious from docs):

| REST (config body) | CLI flag | Gotcha |
|--------------------|----------|--------|
| `"type": "r2_data_catalog"` | `--type r2-data-catalog` | underscores vs hyphens |
| `"table_name"` | `--table` | different key |
| `"token"` | `--catalog-token` | different key |
| `"format": {"type": "parquet"}` | (implied) | required in REST, omitted in CLI |

## Worker Binding

```jsonc
// wrangler.jsonc
{ "pipelines": [ { "stream": "<STREAM_ID>", "binding": "MY_STREAM" } ] }
```

> Binding field is `"stream"` as of June 2026 (was `"pipeline"`, still accepted). Use the **stream ID** (`wrangler pipelines streams list`), not the pipeline ID. Redeploy after adding. Generate typed bindings with `npx wrangler types` → `Pipeline<Cloudflare.MyStreamRecord>` from `cloudflare:pipelines`.

## Terraform

Resources: `cloudflare_pipeline_stream`, `cloudflare_pipeline_sink`, `cloudflare_pipeline`. For current attribute schemas pull `https://developers.cloudflare.com/pipelines/reference/terraform/`.

```hcl
resource "cloudflare_pipeline_stream" "my_stream" {
  account_id     = var.cloudflare_account_id
  name           = "my_stream"
  format         = { type = "json" }
  schema         = { fields = [{ name = "value", type = "json", required = true }] }
  http           = { enabled = true, authentication = false, cors = {} }
  worker_binding = { enabled = false }
}

resource "cloudflare_pipeline_sink" "my_sink" {
  account_id = var.cloudflare_account_id
  name       = "my_sink"
  type       = "r2_data_catalog"
  format     = { type = "parquet" }
  schema     = { fields = [] }
  config     = {
    account_id = var.cloudflare_account_id
    bucket     = cloudflare_r2_bucket.pipeline_bucket.name
    table_name = "my_table"
    token      = var.catalog_token
  }
}

resource "cloudflare_pipeline" "my_pipeline" {
  account_id = var.cloudflare_account_id
  name       = "my_pipeline"
  sql        = "INSERT INTO ${cloudflare_pipeline_sink.my_sink.name} SELECT * FROM ${cloudflare_pipeline_stream.my_stream.name}"
}
```

## Credentials

| Type | Permission |
|------|------------|
| Catalog token (Iceberg sink) | R2 Storage Admin R&W + R2 Data Catalog R&W |
| R2 credentials (raw sink) | Object Read & Write |
| HTTP ingest token | Workers Pipelines Send (only if stream auth enabled) |

## See Also

- [api.md](api.md) — sending events, REST API, lifecycle · [gotchas.md](gotchas.md) — immutability, REST≠CLI

# Pipelines Gotchas

Non-obvious failure modes (not well covered by docs). For current limits and error semantics, pull `https://developers.cloudflare.com/pipelines/platform/limits/`.

## Events accepted but never appear (most common)

HTTP 200 / `send()` resolves, but no data in the sink. Causes:

1. **Schema validation failure** — structured streams accept then **silently drop** invalid events during processing. Validate client-side (Zod) and monitor `pipelinesUserErrorsAdaptiveGroups`.
2. **First-flush warm-up** — first data takes **3–7 minutes** (warm-up + namespace/table creation) even with `--roll-interval 10`. Poll ≥5 min in tests.
3. **Roll interval not elapsed** — default 300s.
4. **Silent sink failure** — deleted bucket or expired token. Check `recordsWritten > 0` but `filesWritten = 0`; inspect `failure_reason` via `GET /pipelines/{id}`.

## Everything is immutable

Cannot modify stream schema, pipeline SQL, or sink config — delete and recreate. Use version naming (`events_v1`) and keep SQL in version control.

```bash
curl -X DELETE "$BASE_URL/pipelines/{id}" -H "Authorization: Bearer $API_TOKEN"
curl -X DELETE "$BASE_URL/sinks/{id}"     -H "Authorization: Bearer $API_TOKEN"
curl -X DELETE "$BASE_URL/streams/{id}"   -H "Authorization: Bearer $API_TOKEN"
```

## Worker binding undefined (`env.MY_STREAM`)

1. Use the **stream ID**, not pipeline ID, in `wrangler.jsonc`.
2. Binding field is `"stream"` (June 2026); old `"pipeline"` still works.
3. Redeploy after adding the binding.

## REST API field names ≠ CLI flags

`r2_data_catalog` vs `--type r2-data-catalog`, `table_name` vs `--table`, `token` vs `--catalog-token`, and `format` is required in REST but implied in CLI. See [configuration.md](configuration.md#option-c-rest-api-programmatic).

## `wrangler pipelines delete` defaults to "no"

Non-interactive environments answer "no" automatically — use REST `DELETE` for CI/automation.

## Behavioral Notes

- **`__ingest_ts` auto-added** (TIMESTAMP, day-partitioned). Don't put it in your schema.
- **Sinks can't target existing tables** — the sink creates its own. Use PySpark to write to existing tables.
- **JSON-only input** — no Avro/Protobuf/CSV.
- **Naming:** streams/sinks/pipelines use underscores; buckets use hyphens.
- **Metrics lag 5–10 min** after creation.
- **Pipeline SQL is row-level only** — no GROUP BY/aggregation/window functions (do aggregation in [R2 SQL](../r2-sql/) at query time). CTEs and `UNNEST` are supported.

## Debug Checklist

- [ ] Stream exists: `wrangler pipelines streams list`
- [ ] Pipeline `running` (not `initializing`/`failed`): `GET /pipelines/{id}`, check `failure_reason`
- [ ] SQL matches schema; sink token valid; bucket + catalog exist
- [ ] Worker redeployed; binding uses **stream ID** under `"stream"`
- [ ] Waited ≥5 min (first flush)
- [ ] Sink metrics: `filesWritten > 0`; error metrics show no drops

## See Also

- [configuration.md](configuration.md) · [api.md](api.md) · [patterns.md](patterns.md)

# Pipelines Patterns

Code-first patterns. For observability dataset/field schemas and Logpush dataset lists, pull `https://developers.cloudflare.com/pipelines/observability/metrics/` and `https://developers.cloudflare.com/pipelines/streams/logpush/`.

## Fire-and-Forget Producer

```typescript
export default {
  async fetch(req, env, ctx) {
    const event = { event_id: crypto.randomUUID(), event_type: "page_view", timestamp: new Date().toISOString() };
    ctx.waitUntil(env.MY_STREAM.send([event]));  // don't block the response
    return new Response("OK");
  }
};
```

## Client-Side Validation with Zod

Structured streams drop invalid events silently during processing. Validate before sending for immediate feedback.

```typescript
import { z } from "zod";

const EventSchema = z.object({
  event_id: z.string(),
  category: z.enum(["purchase", "view"]),
  amount: z.number().positive().optional(),
});

const validated = EventSchema.parse(rawEvent);  // throws synchronously
await env.MY_STREAM.send([validated]);
```

## Scheduled Collector Worker

```jsonc
// wrangler.jsonc
{
  "name": "collector",
  "pipelines": [{ "stream": "<STREAM_ID>", "binding": "EVENT_STREAM" }],
  "triggers": { "crons": ["*/5 * * * *"] }
}
```

```typescript
export default {
  async scheduled(event, env, ctx) {
    const items = await (await fetch("https://api.example.com/data")).json();
    const events = items.map(i => ({
      event_id: crypto.randomUUID(),
      timestamp: new Date().toISOString(),
      category: i.type, amount: i.value,
    }));
    await env.EVENT_STREAM.send(events);
  },
};
```

## Logpush → Pipelines

Pipelines is a native Logpush destination — ingest Cloudflare logs, transform with SQL, store as Iceberg/Parquet. For the current supported dataset list and field names, pull the Logpush doc above.

```sql
INSERT INTO http_logs_sink
SELECT
  ClientIP,
  EdgeResponseStatus,
  to_timestamp_micros(EdgeStartTimestamp) AS event_time,
  upper(ClientRequestMethod) AS method,
  sha256(ClientIP) AS hashed_ip          -- redact PII at ingest
FROM http_logs_stream
WHERE EdgeResponseStatus >= 400;
```

Configure via Dashboard (**Logpush → Create a job → Pipelines** destination) or API.

## Pipelines + Queues Fan-out

```typescript
await Promise.all([
  env.ANALYTICS_STREAM.send([event]),  // long-term storage + SQL
  env.PROCESS_QUEUE.send(event),       // immediate processing + retries
]);
```

Use Pipelines for long-term storage + SQL; Queues for immediate processing/retries/DLQ; both for fan-out.

## Observability (GraphQL Analytics)

Same R2 API token works. Endpoint: `https://api.cloudflare.com/client/v4/graphql`. Datasets cover ingestion, processing (incl. `decodeErrors`), delivery, sink writes (`filesWritten`), and user/validation errors — see the metrics doc for the full dataset/field catalog.

```bash
curl -X POST "https://api.cloudflare.com/client/v4/graphql" \
  -H "Authorization: Bearer $API_TOKEN" -H "Content-Type: application/json" \
  -d '{"query": "query { viewer { accounts(filter: {accountTag: \"'$ACCOUNT_ID'\"}) { pipelinesIngestionAdaptiveGroups(filter: {pipelineId: \"PIPELINE-UUID-WITH-DASHES\", datetime_geq: \"2026-03-01T00:00:00Z\"}, limit: 10) { sum { ingestedRecords ingestedBytes } dimensions { datetimeHour } } } } }"}'
```

> **Sink/pipeline IDs need dashes for GraphQL** but wrangler may show them without: `b909fe6e544844abbd63f6dcbc81d602` → `b909fe6e-5448-44ab-bd63-f6dcbc81d602`. Metrics take 5–10 min to populate.

### Detecting Silent Data Loss

If a sink's bucket is deleted or its token expires, events are accepted but lost. Tell-tale: `recordsWritten > 0` but `filesWritten = 0`. Always verify data lands in R2 within the roll interval and R2 SQL returns expected counts.

## Schema Evolution (Immutable Pipelines)

Pipelines can't change. Version + dual-write:

```bash
npx wrangler pipelines streams create events_v2 --schema-file v2.json
```
```typescript
await Promise.all([env.EVENTS_V1.send([event]), env.EVENTS_V2.send([event])]);
// query across versions with UNION ALL in R2 SQL
```

## End-to-End: Streaming Analytics Dashboard

```
External APIs → Collector Worker (cron) → Pipeline → R2 (Iceberg) → Dashboard Worker → R2 SQL
```

1. Create bucket + enable catalog ([r2-data-catalog](../r2-data-catalog/configuration.md))
2. Create stream + sink + pipeline (here)
3. Collector Worker with cron + stream binding (above)
4. Dashboard Worker querying R2 SQL ([r2-sql/patterns.md](../r2-sql/patterns.md))
5. Enable automatic compaction

## See Also

- [configuration.md](configuration.md) · [api.md](api.md) · [gotchas.md](gotchas.md) · [r2-sql](../r2-sql/)

# Cloudflare Pipelines

Streaming ingest: receive events over HTTP/Workers/Logpush, transform with SQL, write to R2 as Iceberg tables or Parquet/JSON files.

## Documentation

This reference is a fast-start with verified code and gotchas. For limits, settings, full SQL syntax, and pricing, **retrieve the live docs** — use the Cloudflare MCP `docs` tool if available, otherwise `webfetch` the URL. Docs are source of truth over this file.

| Topic | URL |
|-------|-----|
| Overview / getting started | `https://developers.cloudflare.com/pipelines/getting-started/` |
| Streams (write, manage, Logpush) | `https://developers.cloudflare.com/pipelines/streams/` |
| Sinks | `https://developers.cloudflare.com/pipelines/sinks/` |
| Pipelines & SQL transforms | `https://developers.cloudflare.com/pipelines/pipelines/` |
| SQL reference (statements, types) | `https://developers.cloudflare.com/pipelines/sql-reference/` |
| Wrangler commands | `https://developers.cloudflare.com/pipelines/reference/wrangler-commands/` |
| Terraform | `https://developers.cloudflare.com/pipelines/reference/terraform/` |
| Limits | `https://developers.cloudflare.com/pipelines/platform/limits/` |
| Pricing | `https://developers.cloudflare.com/pipelines/platform/pricing/` |
| Metrics (GraphQL) | `https://developers.cloudflare.com/pipelines/observability/metrics/` |

## Three Components

```
Sources → Stream → Pipeline (SQL) → Sink → R2
          ↑          ↓                 ↓
   HTTP / Workers / Transform     Iceberg (Data Catalog)
   Logpush          (row-level)   or Parquet/JSON files
```

| Component | Purpose |
|-----------|---------|
| **Stream** | Receives events (HTTP endpoint, Worker binding, or Logpush). Structured (schema-validated) or unstructured. |
| **Pipeline** | SQL connecting a stream to a sink. Row-level transforms only — no GROUP BY/aggregation. |
| **Sink** | Writes to R2 — Iceberg via Data Catalog, or raw Parquet/JSON. |

**Status:** Open beta (Workers Paid for production). Pricing announced; verify billing status in docs.

## Quick Start

```bash
# Interactive — creates stream + sink + pipeline, optionally bucket + catalog
npx wrangler pipelines setup
```

Minimal Worker producer:
```typescript
interface Env { MY_STREAM: Pipeline; }

export default {
  async fetch(req: Request, env: Env, ctx: ExecutionContext): Promise<Response> {
    ctx.waitUntil(env.MY_STREAM.send([{ event_id: crypto.randomUUID(), amount: 29.99 }]));
    return new Response("OK");
  }
} satisfies ExportedHandler<Env>;
```

## Which Sink Type?

```
Need SQL queries / ACID / time-travel on the data?
  → R2 Data Catalog (Iceberg)   ✅ R2 SQL, schema evolution   ❌ more setup

Just archival / external tools (Spark, Athena)?
  → R2 raw files (Parquet/JSON) ✅ simple, partitioned files  ❌ no built-in SQL
```

## Critical Behaviors (read before building)

These are non-obvious and prevent most failures — see [gotchas.md](gotchas.md) for detail.

- **Everything is immutable after creation** — stream schema, pipeline SQL, sink config. To change, delete and recreate.
- **Sinks create their own table** — they cannot target an existing Iceberg table.
- **`__ingest_ts` is added automatically** (TIMESTAMP, partitioned by day). Don't define it in your schema.
- **Data isn't queryable immediately** — first flush takes **3–7 minutes** (warm-up + table creation) even with a short roll interval.
- **Schema validation is deferred** — invalid events are accepted then silently dropped. Monitor via GraphQL error metrics.
- **Binding field renamed `pipeline` → `stream`** (June 2026); old field still accepted.

## Reading Order

1. [configuration.md](configuration.md) — schema, streams, sinks, pipelines (CLI + REST + Terraform), bindings
2. [api.md](api.md) — `send()`, HTTP ingest, REST API, pipeline SQL, lifecycle states
3. [patterns.md](patterns.md) — fire-and-forget, validation, Logpush, observability, end-to-end
4. [gotchas.md](gotchas.md) — silent drops, immutability, REST≠CLI field names

## See Also

- [r2-data-catalog](../r2-data-catalog/) — Iceberg sink destination
- [r2-sql](../r2-sql/) — query the ingested data
- [r2](../r2/) · [queues](../queues/) · [workers](https://developers.cloudflare.com/workers/)

# API & Data Sources

## Outputs and Exports

Export resource identifiers:

```typescript
export const kvId = kv.id;
export const bucketName = bucket.name;
export const workerUrl = worker.subdomain;
export const dbId = db.id;
```

## Resource Dependencies

Implicit dependencies via outputs:

```typescript
const kv = new cloudflare.WorkersKvNamespace("kv", {
    accountId: accountId,
    title: "my-kv",
});

// Worker depends on KV (implicit via kv.id)
const worker = new cloudflare.WorkerScript("worker", {
    accountId: accountId,
    name: "my-worker",
    content: code,
    kvNamespaceBindings: [{name: "MY_KV", namespaceId: kv.id}], // Creates dependency
});
```

Explicit dependencies:

```typescript
const migration = new command.local.Command("migration", {
    create: pulumi.interpolate`wrangler d1 execute ${db.name} --file ./schema.sql`,
}, {dependsOn: [db]});

const worker = new cloudflare.WorkerScript("worker", {
    accountId: accountId,
    name: "worker",
    content: code,
    d1DatabaseBindings: [{name: "DB", databaseId: db.id}],
}, {dependsOn: [migration]}); // Ensure migrations run first
```

## Using Outputs with API Calls

```typescript
const db = new cloudflare.D1Database("db", {accountId, name: "my-db"});

db.id.apply(async (dbId) => {
    const response = await fetch(
        `https://api.cloudflare.com/client/v4/accounts/${accountId}/d1/database/${dbId}/query`,
        {method: "POST", headers: {"Authorization": `Bearer ${apiToken}`, "Content-Type": "application/json"},
         body: JSON.stringify({sql: "CREATE TABLE users (id INT)"})}
    );
    return response.json();
});
```

## Custom Dynamic Providers

For resources not in provider:

```typescript
import * as pulumi from "@pulumi/pulumi";

class D1MigrationProvider implements pulumi.dynamic.ResourceProvider {
    async create(inputs: any): Promise<pulumi.dynamic.CreateResult> {
        const response = await fetch(
            `https://api.cloudflare.com/client/v4/accounts/${inputs.accountId}/d1/database/${inputs.databaseId}/query`,
            {method: "POST", headers: {"Authorization": `Bearer ${inputs.apiToken}`, "Content-Type": "application/json"},
             body: JSON.stringify({sql: inputs.sql})}
        );
        return {id: `${inputs.databaseId}-${Date.now()}`, outs: await response.json()};
    }
    async update(id: string, olds: any, news: any): Promise<pulumi.dynamic.UpdateResult> {
        if (olds.sql !== news.sql) await this.create(news);
        return {};
    }
    async delete(id: string, props: any): Promise<void> {}
}

class D1Migration extends pulumi.dynamic.Resource {
    constructor(name: string, args: any, opts?: pulumi.CustomResourceOptions) {
        super(new D1MigrationProvider(), name, args, opts);
    }
}

const migration = new D1Migration("migration", {
    accountId, databaseId: db.id, apiToken, sql: "CREATE TABLE users (id INT)",
}, {dependsOn: [db]});
```

## Data Sources

**Get Zone:**
```typescript
const zone = cloudflare.getZone({name: "example.com"});
const zoneId = zone.then(z => z.id);
```

**Get Accounts (via API):**
Use Cloudflare API directly or custom dynamic resources.

## Import Existing Resources

```bash
# Import worker
pulumi import cloudflare:index/workerScript:WorkerScript my-worker <account_id>/<worker_name>

# Import KV namespace
pulumi import cloudflare:index/workersKvNamespace:WorkersKvNamespace my-kv <namespace_id>

# Import R2 bucket
pulumi import cloudflare:index/r2Bucket:R2Bucket my-bucket <account_id>/<bucket_name>

# Import D1 database
pulumi import cloudflare:index/d1Database:D1Database my-db <account_id>/<database_id>

# Import DNS record
pulumi import cloudflare:index/dnsRecord:DnsRecord my-record <zone_id>/<record_id>
```

## Secrets Management

```typescript
import * as pulumi from "@pulumi/pulumi";

const config = new pulumi.Config();
const apiKey = config.requireSecret("apiKey"); // Encrypted in state

const worker = new cloudflare.WorkerScript("worker", {
    accountId: accountId,
    name: "my-worker",
    content: code,
    secretTextBindings: [{name: "API_KEY", text: apiKey}],
});
```

Store secrets:
```bash
pulumi config set --secret apiKey "secret-value"
```

## Transform Pattern

Modify resource args before creation:

```typescript
import {Transform} from "@pulumi/pulumi";

interface BucketArgs {
    accountId: pulumi.Input<string>;
    transform?: {bucket?: Transform<cloudflare.R2BucketArgs>};
}

function createBucket(name: string, args: BucketArgs) {
    const bucketArgs: cloudflare.R2BucketArgs = {
        accountId: args.accountId,
        name: name,
        location: "auto",
    };
    const finalArgs = args.transform?.bucket?.(bucketArgs) ?? bucketArgs;
    return new cloudflare.R2Bucket(name, finalArgs);
}
```

## v6.x Worker Versioning Resources

**Worker** - Container for versions:
```typescript
const worker = new cloudflare.Worker("api", {accountId, name: "api-worker"});
export const workerId = worker.id;
```

**WorkerVersion** - Immutable code + config:
```typescript
const version = new cloudflare.WorkerVersion("v1", {
    accountId, workerId: worker.id,
    content: fs.readFileSync("./dist/worker.js", "utf8"),
    compatibilityDate: "2025-01-01",
});
export const versionId = version.id;
```

**WorkersDeployment** - Active deployment with bindings:
```typescript
const deployment = new cloudflare.WorkersDeployment("prod", {
    accountId, workerId: worker.id, versionId: version.id,
    kvNamespaceBindings: [{name: "MY_KV", namespaceId: kv.id}],
});
```

**Use:** Advanced deployments (canary, blue-green). Most apps should use `WorkerScript` (auto-versioning).

---
See: [README.md](./README.md), [configuration.md](./configuration.md), [patterns.md](./patterns.md), [gotchas.md](./gotchas.md)

# Resource Configuration

## Workers (cloudflare.WorkerScript)

```typescript
import * as cloudflare from "@pulumi/cloudflare";
import * as fs from "fs";

const worker = new cloudflare.WorkerScript("my-worker", {
    accountId: accountId,
    name: "my-worker",
    content: fs.readFileSync("./dist/worker.js", "utf8"),
    module: true, // ES modules
    compatibilityDate: "2025-01-01",
    compatibilityFlags: ["nodejs_compat"],
    
    // v6.x: Observability
    logpush: true, // Enable Workers Logpush
    tailConsumers: [{service: "log-consumer"}], // Stream logs to Worker
    
    // v6.x: Placement
    placement: {mode: "smart"}, // Smart placement for latency optimization
    
    // Bindings
    kvNamespaceBindings: [{name: "MY_KV", namespaceId: kv.id}],
    r2BucketBindings: [{name: "MY_BUCKET", bucketName: bucket.name}],
    d1DatabaseBindings: [{name: "DB", databaseId: db.id}],
    queueBindings: [{name: "MY_QUEUE", queue: queue.id}],
    serviceBindings: [{name: "OTHER_SERVICE", service: other.name}],
    plainTextBindings: [{name: "ENV_VAR", text: "value"}],
    secretTextBindings: [{name: "API_KEY", text: secret}],
    
    // v6.x: Advanced bindings
    analyticsEngineBindings: [{name: "ANALYTICS", dataset: "my-dataset"}],
    browserBinding: {name: "BROWSER"}, // Browser Rendering
    aiBinding: {name: "AI"}, // Workers AI
    hyperdriveBindings: [{name: "HYPERDRIVE", id: hyperdriveConfig.id}],
});
```

## Workers KV (cloudflare.WorkersKvNamespace)

```typescript
const kv = new cloudflare.WorkersKvNamespace("my-kv", {
    accountId: accountId,
    title: "my-kv-namespace",
});

// Write values
const kvValue = new cloudflare.WorkersKvValue("config", {
    accountId: accountId,
    namespaceId: kv.id,
    key: "config",
    value: JSON.stringify({foo: "bar"}),
});
```

## R2 Buckets (cloudflare.R2Bucket)

```typescript
const bucket = new cloudflare.R2Bucket("my-bucket", {
    accountId: accountId,
    name: "my-bucket",
    location: "auto", // or "wnam", etc.
});
```

## D1 Databases (cloudflare.D1Database)

```typescript
const db = new cloudflare.D1Database("my-db", {accountId, name: "my-database"});

// Migrations via wrangler
import * as command from "@pulumi/command";
const migration = new command.local.Command("d1-migration", {
    create: pulumi.interpolate`wrangler d1 execute ${db.name} --file ./schema.sql`,
}, {dependsOn: [db]});
```

## Queues (cloudflare.Queue)

```typescript
const queue = new cloudflare.Queue("my-queue", {accountId, name: "my-queue"});

// Producer
const producer = new cloudflare.WorkerScript("producer", {
    accountId, name: "producer", content: code,
    queueBindings: [{name: "MY_QUEUE", queue: queue.id}],
});

// Consumer
const consumer = new cloudflare.WorkerScript("consumer", {
    accountId, name: "consumer", content: code,
    queueConsumers: [{queue: queue.name, maxBatchSize: 10, maxRetries: 3}],
});
```

## Pages Projects (cloudflare.PagesProject)

```typescript
const pages = new cloudflare.PagesProject("my-site", {
    accountId, name: "my-site", productionBranch: "main",
    buildConfig: {buildCommand: "npm run build", destinationDir: "dist"},
    source: {
        type: "github",
        config: {owner: "my-org", repoName: "my-repo", productionBranch: "main"},
    },
    deploymentConfigs: {
        production: {
            environmentVariables: {NODE_VERSION: "18"},
            kvNamespaces: {MY_KV: kv.id},
            d1Databases: {DB: db.id},
        },
    },
});
```

## DNS Records (cloudflare.DnsRecord)

```typescript
const zone = cloudflare.getZone({name: "example.com"});
const record = new cloudflare.DnsRecord("www", {
    zoneId: zone.then(z => z.id), name: "www", type: "A",
    content: "192.0.2.1", ttl: 3600, proxied: true,
});
```

## Workers Domains/Routes

```typescript
// Route (pattern-based)
const route = new cloudflare.WorkerRoute("my-route", {
    zoneId: zoneId,
    pattern: "example.com/api/*",
    scriptName: worker.name,
});

// Domain (dedicated subdomain)
const domain = new cloudflare.WorkersDomain("my-domain", {
    accountId: accountId,
    hostname: "api.example.com",
    service: worker.name,
    zoneId: zoneId,
});
```

## Assets Configuration (v6.x)

Serve static assets from Workers:

```typescript
const worker = new cloudflare.WorkerScript("app", {
    accountId: accountId,
    name: "my-app",
    content: code,
    assets: {
        path: "./public", // Local directory
        // Assets uploaded and served from Workers
    },
});
```

## v6.x Versioned Deployments (Advanced)

For gradual rollouts, use 3-resource pattern:

```typescript
// 1. Worker (container for versions)
const worker = new cloudflare.Worker("api", {
    accountId: accountId,
    name: "api-worker",
});

// 2. Version (immutable code + config)
const version = new cloudflare.WorkerVersion("v1", {
    accountId: accountId,
    workerId: worker.id,
    content: fs.readFileSync("./dist/worker.js", "utf8"),
    compatibilityDate: "2025-01-01",
    compatibilityFlags: ["nodejs_compat"],
    // Note: Bindings configured at deployment level
});

// 3. Deployment (version + bindings + traffic split)
const deployment = new cloudflare.WorkersDeployment("prod", {
    accountId: accountId,
    workerId: worker.id,
    versionId: version.id,
    // Bindings applied to deployment
    kvNamespaceBindings: [{name: "MY_KV", namespaceId: kv.id}],
});
```

**When to use:** Blue-green deployments, canary releases, gradual rollouts  
**When NOT to use:** Simple single-version deployments (use WorkerScript)

---
See: [README.md](./README.md), [api.md](./api.md), [patterns.md](./patterns.md), [gotchas.md](./gotchas.md)

# Troubleshooting & Best Practices

## Common Errors

### "No bundler/build step" - Pulumi uploads raw code

**Problem:** Worker fails with "Cannot use import statement outside a module"  
**Cause:** Pulumi doesn't bundle Worker code - uploads exactly what you provide  
**Solution:** Build Worker BEFORE Pulumi deploy

```typescript
// WRONG: Pulumi won't bundle this
const worker = new cloudflare.WorkerScript("worker", {
    content: fs.readFileSync("./src/index.ts", "utf8"), // Raw TS file
});

// RIGHT: Build first, then deploy
import * as command from "@pulumi/command";
const build = new command.local.Command("build", {
    create: "npm run build",
    dir: "./worker",
});
const worker = new cloudflare.WorkerScript("worker", {
    content: build.stdout.apply(() => fs.readFileSync("./worker/dist/index.js", "utf8")),
}, {dependsOn: [build]});
```

### "wrangler.toml not consumed" - Config drift

**Problem:** Local wrangler dev works, Pulumi deploy fails  
**Cause:** Pulumi ignores wrangler.toml - must duplicate config  
**Solution:** Generate wrangler.toml from Pulumi or keep synced manually

```typescript
// Pattern: Export Pulumi config to wrangler.toml
const workerConfig = {
    name: "my-worker",
    compatibilityDate: "2025-01-01",
    compatibilityFlags: ["nodejs_compat"],
};

new command.local.Command("generate-wrangler", {
    create: pulumi.interpolate`cat > wrangler.toml <<EOF
name = "${workerConfig.name}"
compatibility_date = "${workerConfig.compatibilityDate}"
compatibility_flags = ${JSON.stringify(workerConfig.compatibilityFlags)}
EOF`,
});
```

### "False no-changes detection" - Content SHA unchanged

**Problem:** Worker code updated, Pulumi says "no changes"  
**Cause:** Content hash identical (whitespace/comment-only change)  
**Solution:** Add build timestamp or version to force update

```typescript
const version = Date.now().toString();
const worker = new cloudflare.WorkerScript("worker", {
    content: code,
    plainTextBindings: [{name: "VERSION", text: version}], // Forces new deployment
});
```

### "D1 migrations don't run on pulumi up"

**Problem:** Database schema not applied after D1 database created  
**Cause:** Pulumi creates database but doesn't run migrations  
**Solution:** Use Command resource with dependsOn

```typescript
const db = new cloudflare.D1Database("db", {accountId, name: "mydb"});

// Run migrations after DB created
const migration = new command.local.Command("migrate", {
    create: pulumi.interpolate`wrangler d1 execute ${db.name} --file ./schema.sql`,
}, {dependsOn: [db]});

// Worker depends on migrations
const worker = new cloudflare.WorkerScript("worker", {
    d1DatabaseBindings: [{name: "DB", databaseId: db.id}],
}, {dependsOn: [migration]});
```

### "Missing required property 'accountId'"

**Problem:** `Error: Missing required property 'accountId'`  
**Cause:** Account ID not provided in resource configuration  
**Solution:** Add to stack config

```yaml
# Pulumi.<stack>.yaml
config:
  cloudflare:accountId: "abc123..."
```

### "Binding name mismatch"

**Problem:** Worker fails with "env.MY_KV is undefined"  
**Cause:** Binding name in Pulumi != name in Worker code  
**Solution:** Match exactly (case-sensitive)

```typescript
// Pulumi
kvNamespaceBindings: [{name: "MY_KV", namespaceId: kv.id}]

// Worker code
export default { async fetch(request, env) { await env.MY_KV.get("key"); }}
```

### "API token permissions insufficient"

**Problem:** `Error: authentication error (10000)`  
**Cause:** Token lacks required permissions  
**Solution:** Grant token permissions: Account.Workers Scripts:Edit, Account.Account Settings:Read

### "Resource not found after import"

**Problem:** Imported resource shows as changed on next `pulumi up`  
**Cause:** State mismatch between actual resource and Pulumi config  
**Solution:** Check property names/types match exactly

```bash
pulumi import cloudflare:index/workerScript:WorkerScript my-worker <account_id>/<worker_name>
pulumi preview # If shows changes, adjust Pulumi code to match actual resource
```

### "v6.x Worker versioning confusion"

**Problem:** Worker deployed but not receiving traffic  
**Cause:** v6.x requires Worker + WorkerVersion + WorkersDeployment (3 resources)  
**Solution:** Use WorkerScript (auto-versioning) OR full versioning pattern

```typescript
// SIMPLE: WorkerScript auto-versions (default behavior)
const worker = new cloudflare.WorkerScript("worker", {
    accountId, name: "my-worker", content: code,
});

// ADVANCED: Manual versioning for gradual rollouts (v6.x)
const worker = new cloudflare.Worker("worker", {accountId, name: "my-worker"});
const version = new cloudflare.WorkerVersion("v1", {
    accountId, workerId: worker.id, content: code, compatibilityDate: "2025-01-01",
});
const deployment = new cloudflare.WorkersDeployment("prod", {
    accountId, workerId: worker.id, versionId: version.id,
});
```

## Best Practices

1. **Always set compatibilityDate** - Locks Worker behavior, prevents breaking changes
2. **Build before deploy** - Pulumi doesn't bundle; use Command resource or CI build step
3. **Match binding names** - Case-sensitive, must match between Pulumi and Worker code
4. **Use dependsOn for migrations** - Ensure D1 migrations run before Worker deploys
5. **Version Worker content** - Add VERSION binding to force redeployment on content changes
6. **Store secrets in stack config** - Use `pulumi config set --secret` for API keys

## Limits

| Resource | Limit | Notes |
|----------|-------|-------|
| Worker script size | 10 MB | Includes all dependencies, after compression |
| Worker CPU time | 10ms (free), 30s default / 5min max (paid) | Per request |
| KV keys per namespace | Unlimited | 1000 ops/sec write, 100k ops/sec read |
| R2 storage | Unlimited | Class A ops: 1M/mo free, Class B: 10M/mo free |
| D1 databases | 50,000 per account | Free: 10 per account, 5 GB each |
| Queues | 10,000 per account | Free: 1M ops/day |
| Pages projects | 500 per account | Free: 100 projects |
| API requests | Varies by plan | ~1200 req/5min on free |

## Resources

- **Pulumi Registry:** https://www.pulumi.com/registry/packages/cloudflare/
- **API Docs:** https://www.pulumi.com/registry/packages/cloudflare/api-docs/
- **GitHub:** https://github.com/pulumi/pulumi-cloudflare
- **Cloudflare Docs:** https://developers.cloudflare.com/
- **Workers Docs:** https://developers.cloudflare.com/workers/

---
See: [README.md](./README.md), [configuration.md](./configuration.md), [api.md](./api.md), [patterns.md](./patterns.md)

# Architecture Patterns

## Component Resources

```typescript
class WorkerApp extends pulumi.ComponentResource {
    constructor(name: string, args: WorkerAppArgs, opts?) {
        super("custom:cloudflare:WorkerApp", name, {}, opts);
        const defaultOpts = {parent: this};

        this.kv = new cloudflare.WorkersKvNamespace(`${name}-kv`, {accountId: args.accountId, title: `${name}-kv`}, defaultOpts);
        this.worker = new cloudflare.WorkerScript(`${name}-worker`, {
            accountId: args.accountId, name: `${name}-worker`, content: args.workerCode,
            module: true, kvNamespaceBindings: [{name: "KV", namespaceId: this.kv.id}],
        }, defaultOpts);
        this.domain = new cloudflare.WorkersDomain(`${name}-domain`, {
            accountId: args.accountId, hostname: args.domain, service: this.worker.name,
        }, defaultOpts);
    }
}
```

## Full-Stack Worker App

```typescript
const kv = new cloudflare.WorkersKvNamespace("cache", {accountId, title: "api-cache"});
const db = new cloudflare.D1Database("db", {accountId, name: "app-database"});
const bucket = new cloudflare.R2Bucket("assets", {accountId, name: "app-assets"});

const apiWorker = new cloudflare.WorkerScript("api", {
    accountId, name: "api-worker", content: fs.readFileSync("./dist/api.js", "utf8"),
    module: true, kvNamespaceBindings: [{name: "CACHE", namespaceId: kv.id}],
    d1DatabaseBindings: [{name: "DB", databaseId: db.id}],
    r2BucketBindings: [{name: "ASSETS", bucketName: bucket.name}],
});
```

## Multi-Environment Setup

```typescript
const stack = pulumi.getStack();
const worker = new cloudflare.WorkerScript(`worker-${stack}`, {
    accountId, name: `my-worker-${stack}`, content: code,
    plainTextBindings: [{name: "ENVIRONMENT", text: stack}],
});
```

## Queue-Based Processing

```typescript
const queue = new cloudflare.Queue("processing-queue", {accountId, name: "image-processing"});

// Producer: API receives requests
const apiWorker = new cloudflare.WorkerScript("api", {
    accountId, name: "api-worker", content: apiCode,
    queueBindings: [{name: "PROCESSING_QUEUE", queue: queue.id}],
});

// Consumer: Process async
const processorWorker = new cloudflare.WorkerScript("processor", {
    accountId, name: "processor-worker", content: processorCode,
    queueConsumers: [{queue: queue.name, maxBatchSize: 10, maxRetries: 3, maxWaitTimeMs: 5000}],
    r2BucketBindings: [{name: "OUTPUT_BUCKET", bucketName: outputBucket.name}],
});
```

## Microservices with Service Bindings

```typescript
const authWorker = new cloudflare.WorkerScript("auth", {accountId, name: "auth-service", content: authCode});
const apiWorker = new cloudflare.WorkerScript("api", {
    accountId, name: "api-service", content: apiCode,
    serviceBindings: [{name: "AUTH", service: authWorker.name}],
});
```

## Event-Driven Architecture

```typescript
const eventQueue = new cloudflare.Queue("events", {accountId, name: "event-bus"});
const producer = new cloudflare.WorkerScript("producer", {
    accountId, name: "api-producer", content: producerCode,
    queueBindings: [{name: "EVENTS", queue: eventQueue.id}],
});
const consumer = new cloudflare.WorkerScript("consumer", {
    accountId, name: "email-consumer", content: consumerCode,
    queueConsumers: [{queue: eventQueue.name, maxBatchSize: 10}],
});
```

## v6.x Versioned Deployments (Blue-Green/Canary)

```typescript
const worker = new cloudflare.Worker("api", {accountId, name: "api-worker"});
const v1 = new cloudflare.WorkerVersion("v1", {accountId, workerId: worker.id, content: fs.readFileSync("./dist/v1.js", "utf8"), compatibilityDate: "2025-01-01"});
const v2 = new cloudflare.WorkerVersion("v2", {accountId, workerId: worker.id, content: fs.readFileSync("./dist/v2.js", "utf8"), compatibilityDate: "2025-01-01"});

// Gradual rollout: 10% v2, 90% v1
const deployment = new cloudflare.WorkersDeployment("canary", {
    accountId, workerId: worker.id,
    versions: [{versionId: v2.id, percentage: 10}, {versionId: v1.id, percentage: 90}],
    kvNamespaceBindings: [{name: "MY_KV", namespaceId: kv.id}],
});
```

**Use:** Canary releases, A/B testing, blue-green. Most apps use `WorkerScript` (auto-versioning).

## Wrangler.toml Generation (Bridge IaC with Local Dev)

Generate wrangler.toml from Pulumi config to keep local dev in sync:

```typescript
import * as command from "@pulumi/command";

const workerConfig = {
    name: "my-worker",
    compatibilityDate: "2025-01-01",
    compatibilityFlags: ["nodejs_compat"],
};

// Create resources
const kv = new cloudflare.WorkersKvNamespace("kv", {accountId, title: "my-kv"});
const db = new cloudflare.D1Database("db", {accountId, name: "my-db"});
const bucket = new cloudflare.R2Bucket("bucket", {accountId, name: "my-bucket"});

// Generate wrangler.toml after resources created
const wranglerGen = new command.local.Command("gen-wrangler", {
    create: pulumi.interpolate`cat > wrangler.toml <<EOF
name = "${workerConfig.name}"
main = "src/index.ts"
compatibility_date = "${workerConfig.compatibilityDate}"
compatibility_flags = ${JSON.stringify(workerConfig.compatibilityFlags)}

[[kv_namespaces]]
binding = "MY_KV"
id = "${kv.id}"

[[d1_databases]]
binding = "DB"
database_id = "${db.id}"
database_name = "${db.name}"

[[r2_buckets]]
binding = "MY_BUCKET"
bucket_name = "${bucket.name}"
EOF`,
}, {dependsOn: [kv, db, bucket]});

// Deploy worker after wrangler.toml generated
const worker = new cloudflare.WorkerScript("worker", {
    accountId, name: workerConfig.name, content: code,
    compatibilityDate: workerConfig.compatibilityDate,
    compatibilityFlags: workerConfig.compatibilityFlags,
    kvNamespaceBindings: [{name: "MY_KV", namespaceId: kv.id}],
    d1DatabaseBindings: [{name: "DB", databaseId: db.id}],
    r2BucketBindings: [{name: "MY_BUCKET", bucketName: bucket.name}],
}, {dependsOn: [wranglerGen]});
```

**Benefits:**
- `wrangler dev` uses same bindings as production
- No config drift between Pulumi and local dev
- Single source of truth (Pulumi config)

**Alternative:** Read wrangler.toml in Pulumi (reverse direction) if wrangler is source of truth

## Build + Deploy Pattern

```typescript
import * as command from "@pulumi/command";
const build = new command.local.Command("build", {create: "npm run build", dir: "./worker"});
const worker = new cloudflare.WorkerScript("worker", {
    accountId, name: "my-worker",
    content: build.stdout.apply(() => fs.readFileSync("./worker/dist/index.js", "utf8")),
}, {dependsOn: [build]});
```

## Content SHA Pattern (Force Updates)

Prevent false "no changes" detections:

```typescript
const version = Date.now().toString();
const worker = new cloudflare.WorkerScript("worker", {
    accountId, name: "my-worker", content: code,
    plainTextBindings: [{name: "VERSION", text: version}], // Forces deployment
});
```

---
See: [README.md](./README.md), [configuration.md](./configuration.md), [api.md](./api.md), [gotchas.md](./gotchas.md)

# Cloudflare Pulumi Provider

Expert guidance for Cloudflare Pulumi Provider (@pulumi/cloudflare).

## Overview

Programmatic management of Cloudflare resources: Workers, Pages, D1, KV, R2, DNS, Queues, etc.

**Packages:**
- TypeScript/JS: `@pulumi/cloudflare`
- Python: `pulumi-cloudflare`
- Go: `github.com/pulumi/pulumi-cloudflare/sdk/v6/go/cloudflare`
- .NET: `Pulumi.Cloudflare`

**Version:** v6.x

## Core Principles

1. Use API tokens (not legacy API keys)
2. Store accountId in stack config
3. Match binding names across code/config
4. Use `module: true` for ES modules
5. Set `compatibilityDate` to lock behavior

## Authentication

```typescript
import * as cloudflare from "@pulumi/cloudflare";

// API Token (recommended): CLOUDFLARE_API_TOKEN env
const provider = new cloudflare.Provider("cf", { apiToken: process.env.CLOUDFLARE_API_TOKEN });

// API Key (legacy): CLOUDFLARE_API_KEY + CLOUDFLARE_EMAIL env
const provider = new cloudflare.Provider("cf", { apiKey: process.env.CLOUDFLARE_API_KEY, email: process.env.CLOUDFLARE_EMAIL });

// API User Service Key: CLOUDFLARE_API_USER_SERVICE_KEY env
const provider = new cloudflare.Provider("cf", { apiUserServiceKey: process.env.CLOUDFLARE_API_USER_SERVICE_KEY });
```

## Setup

**Pulumi.yaml:**
```yaml
name: my-cloudflare-app
runtime: nodejs
config:
  cloudflare:apiToken:
    value: ${CLOUDFLARE_API_TOKEN}
```

**Pulumi.<stack>.yaml:**
```yaml
config:
  cloudflare:accountId: "abc123..."
```

**index.ts:**
```typescript
import * as pulumi from "@pulumi/pulumi";
import * as cloudflare from "@pulumi/cloudflare";
const accountId = new pulumi.Config("cloudflare").require("accountId");
```

## Common Resource Types
- `Provider` - Provider config
- `WorkerScript` - Worker
- `WorkersKvNamespace` - KV
- `R2Bucket` - R2
- `D1Database` - D1
- `Queue` - Queue
- `PagesProject` - Pages
- `DnsRecord` - DNS
- `WorkerRoute` - Worker route
- `WorkersDomain` - Custom domain

## Key Properties
- `accountId` - Required for most resources
- `zoneId` - Required for DNS/domain
- `name`/`title` - Resource identifier
- `*Bindings` - Connect resources to Workers

## Reading Order

| Order | File | What | When to Read |
|-------|------|------|--------------|
| 1 | [configuration.md](./configuration.md) | Resource config for Workers/KV/D1/R2/Queues/Pages | First time setup, resource reference |
| 2 | [patterns.md](./patterns.md) | Architecture patterns, multi-env, component resources | Building complex apps, best practices |
| 3 | [api.md](./api.md) | Outputs, dependencies, imports, dynamic providers | Advanced features, integrations |
| 4 | [gotchas.md](./gotchas.md) | Common errors, troubleshooting, limits | Debugging, deployment issues |

## In This Reference
- [configuration.md](./configuration.md) - Provider config, stack setup, Workers/bindings
- [api.md](./api.md) - Resource types, Workers script, KV/D1/R2/queues/Pages
- [patterns.md](./patterns.md) - Multi-env, secrets, CI/CD, stack management
- [gotchas.md](./gotchas.md) - State issues, deployment failures, limits

## See Also
- [terraform](../terraform/) - Alternative IaC for Cloudflare
- [wrangler](https://developers.cloudflare.com/workers/wrangler/) - CLI deployment alternative
- [workers](https://developers.cloudflare.com/workers/) - Worker runtime documentation

# Queues API Reference

Fetch the current API documentation for the operation being implemented; do not infer signatures, payloads, or acknowledgement rules from old examples.

| Task | Documentation |
|------|---------------|
| Send individual messages or batches; choose encoding; implement a typed Worker queue handler; dispatch by queue name | [JavaScript APIs](https://developers.cloudflare.com/queues/configuration/javascript-apis/) |
| Understand automatic acknowledgement, explicit per-message and batch actions, precedence, delivery failures, delays, and backoff | [Batching, retries, and delays](https://developers.cloudflare.com/queues/configuration/batching-retries/) |
| Pull over HTTP and acknowledge or retry using leases | [Pull consumers](https://developers.cloudflare.com/queues/configuration/pull-consumers/) |
| Publish from outside Workers | [Publish to a Queue via HTTP](https://developers.cloudflare.com/queues/examples/publish-to-a-queue-via-http/) |

Acknowledge only after the intended work succeeds. For independently processed messages, use per-message outcomes to avoid replaying successful work when another message fails. If catching an error and continuing, explicitly request a retry for work that still needs processing; a successful handler return can acknowledge messages automatically. Fetch the linked acknowledgement rules before mixing message-level and batch-level actions.

Await required work, including downstream writes or sends, before acknowledging it. Check the JavaScript API's handler lifecycle rules before using `waitUntil()`; background work is not independent of delivery success.

See [configuration.md](./configuration.md) for bindings and consumer setup, and [gotchas.md](./gotchas.md) for delivery diagnostics.

# Queues Configuration

Fetch the relevant guide before writing configuration or running CLI commands. Check the project's Wrangler version and compatibility date when adapting examples.

| Task | Documentation |
|------|---------------|
| Create a queue and connect producer and consumer Workers | [Getting started](https://developers.cloudflare.com/queues/get-started/) |
| Configure producer bindings, Worker consumers, retention, and concurrency settings | [Configure Queues](https://developers.cloudflare.com/queues/configuration/configure-queues/) |
| Configure an external HTTP consumer and its visibility timeout | [Pull consumers](https://developers.cloudflare.com/queues/configuration/pull-consumers/) |
| Choose batching, retry policy, or delivery delays | [Batching, retries, and delays](https://developers.cloudflare.com/queues/configuration/batching-retries/) |
| Preserve messages that exhaust retries | [Dead Letter Queues](https://developers.cloudflare.com/queues/configuration/dead-letter-queues/) |
| Set consumer scaling for downstream capacity | [Consumer concurrency](https://developers.cloudflare.com/queues/configuration/consumer-concurrency/) |
| Choose content types and type Worker messages | [JavaScript APIs](https://developers.cloudflare.com/queues/configuration/javascript-apis/) |
| Create, update, attach, remove, or delete queues and consumers | [Wrangler commands](https://developers.cloudflare.com/queues/reference/wrangler-commands/) |
| Pause delivery, resume it, or purge messages | [Pause and purge](https://developers.cloudflare.com/queues/configuration/pause-purge/) |
| Develop and test producers and consumers locally | [Local development](https://developers.cloudflare.com/queues/configuration/local-development/) |

Choose push or pull based on where processing runs, then select an encoding supported by that consumer. Tune batching for acceptable latency and downstream write capacity. Decide how failed messages will be inspected and replayed before configuring a dead-letter queue.

Fetch [limits](https://developers.cloudflare.com/queues/platform/limits/) and [pricing](https://developers.cloudflare.com/queues/platform/pricing/) for the account's plan before selecting retention, delays, or capacity. Do not reuse numeric settings from unrelated examples.

# Queues Gotchas & Troubleshooting

Fetch the linked documentation before changing retry policy or interpreting delivery behavior.

| Symptom or question | Documentation and decision |
|---------------------|----------------------------|
| Successful work repeats after another message fails | Read [acknowledgement and retry rules](https://developers.cloudflare.com/queues/configuration/batching-retries/); use per-message outcomes for independent work. |
| A caught failure disappears instead of retrying | Read [handler lifecycle and APIs](https://developers.cloudflare.com/queues/configuration/javascript-apis/); returning successfully can acknowledge messages. Explicitly retry failed work when continuing. |
| Duplicate processing | Read [delivery guarantees](https://developers.cloudflare.com/queues/reference/delivery-guarantees/); enforce idempotency at the side-effect destination. |
| Pull consumers cannot decode payloads | Check [pull consumer encoding](https://developers.cloudflare.com/queues/configuration/pull-consumers/) and [content types](https://developers.cloudflare.com/queues/configuration/javascript-apis/) against the producer. |
| Messages stop arriving or backlog grows | Check [consumer configuration](https://developers.cloudflare.com/queues/configuration/configure-queues/), [pause state](https://developers.cloudflare.com/queues/configuration/pause-purge/), and [queue metrics](https://developers.cloudflare.com/queues/observability/metrics/). |
| Dead-letter volume rises or messages disappear after retries | Read [Dead Letter Queues](https://developers.cloudflare.com/queues/configuration/dead-letter-queues/); inspect failures and plan recovery before increasing retries. |
| API errors, resource exhaustion, or CPU failures | Read [error codes](https://developers.cloudflare.com/queues/reference/error-codes/), [limits](https://developers.cloudflare.com/queues/platform/limits/), and [consumer concurrency](https://developers.cloudflare.com/queues/configuration/consumer-concurrency/). |
| Retention, delay, throughput, or cost assumptions no longer hold | Retrieve current [limits](https://developers.cloudflare.com/queues/platform/limits/) and [pricing](https://developers.cloudflare.com/queues/platform/pricing/) for the account's plan. |

Distinguish transient dependency failures from invalid payloads before choosing retry or recovery behavior. Acknowledging a failed message does not send it to a dead-letter queue. If handling a permanent failure separately, persist the intended recovery record successfully before acknowledging; use the documented dead-letter policy when relying on retry exhaustion.

See [patterns.md](./patterns.md) for idempotency and downstream integration decisions.

# Queues Patterns & Best Practices

Fetch the guide matching the task and adapt its example to the application's delivery and failure requirements.

| Task | Documentation |
|------|---------------|
| Accept requests and enqueue asynchronous tasks; publish to multiple queues | [Publish to a Queue via Workers](https://developers.cloudflare.com/queues/examples/publish-to-a-queue-via-workers/) |
| Buffer writes to an external API or defer a job | [Batching, retries, and delays](https://developers.cloudflare.com/queues/configuration/batching-retries/) |
| Handle upstream rate limits and backpressure | [Handle rate limits of external APIs](https://developers.cloudflare.com/queues/tutorials/handle-rate-limits/) and [consumer concurrency](https://developers.cloudflare.com/queues/configuration/consumer-concurrency/) |
| Isolate workloads with different latency or capacity needs | [Configure Queues](https://developers.cloudflare.com/queues/configuration/configure-queues/) |
| Retain exhausted retries for inspection and recovery | [Dead Letter Queues](https://developers.cloudflare.com/queues/configuration/dead-letter-queues/) |
| Process R2 object events | [R2 event notifications](https://developers.cloudflare.com/r2/buckets/event-notifications/) |
| Batch output into R2 | [Use Queues to store data in R2](https://developers.cloudflare.com/queues/examples/send-errors-to-r2/) |
| Batch writes into D1 | [D1 database API](https://developers.cloudflare.com/d1/worker-api/d1-database/) |
| Start durable multi-step jobs | [Trigger Workflows](https://developers.cloudflare.com/workflows/build/trigger-workflows/) |
| Publish from a Durable Object | [Use Queues from Durable Objects](https://developers.cloudflare.com/queues/examples/use-queues-with-durable-objects/) |
| Route consumer work to a Durable Object | [Invoke Durable Object methods](https://developers.cloudflare.com/durable-objects/best-practices/create-durable-object-stubs-and-send-requests/) |

Design side effects for [at-least-once delivery](https://developers.cloudflare.com/queues/reference/delivery-guarantees/). A separate check-then-write deduplication flag is not an atomic guarantee: concurrent delivery or a crash between the side effect and recording completion can repeat work. Prefer idempotency keys or transactional enforcement at the destination.

Acknowledge after the destination confirms success. For fan-out, plan for some sends succeeding before another fails; retries must not duplicate downstream effects. Separate queues can isolate workloads, but do not imply a global priority or ordering guarantee.

# Cloudflare Queues

Use Queues to decouple producers from asynchronous consumers and buffer bursts of work. Design consumers for duplicate delivery; use Workflows when the task needs durable multi-step orchestration.

Fetch the relevant documentation below before implementing. Treat current Cloudflare docs as the source of truth for API signatures, acknowledgement semantics, configuration, limits, and pricing.

## Choose a consumer

- Use a Worker push consumer when processing runs on Workers.
- Use an HTTP pull consumer when processing runs in another environment; plan for polling, visibility timeouts, and acknowledgement leases.
- Choose a message encoding the consumer can decode. Check serialization and compatibility-date behavior before sending existing application objects.

See [How Queues works](https://developers.cloudflare.com/queues/reference/how-queues-works/) and [delivery guarantees](https://developers.cloudflare.com/queues/reference/delivery-guarantees/) before choosing ordering or deduplication strategies.

## Read by task

| Task | Reference |
|------|-----------|
| Create queues, bind producers, and configure consumers | [configuration.md](./configuration.md) |
| Send messages and implement acknowledgement or retries | [api.md](./api.md) |
| Buffer APIs, defer jobs, or integrate with storage and orchestration | [patterns.md](./patterns.md) |
| Diagnose delivery failures, duplicates, or capacity issues | [gotchas.md](./gotchas.md) |

For a first application, fetch [Getting started](https://developers.cloudflare.com/queues/get-started/). Retrieve [limits](https://developers.cloudflare.com/queues/platform/limits/) and [pricing](https://developers.cloudflare.com/queues/platform/pricing/) before sizing throughput, retention, or cost; plan-specific values are not maintained here.

# R2 API Reference

Fetch the relevant page before writing code. Use the Workers API for bucket bindings and the S3 API for S3 clients; their types and semantics differ.

| Task | Current documentation |
|------|-----------------------|
| Read, write, inspect, delete, or list objects; metadata, checksums, ranges, and return types | [Workers API reference](https://developers.cloudflare.com/r2/api/workers/workers-api-reference/) |
| Implement a Worker that serves or writes objects | [Use R2 from Workers](https://developers.cloudflare.com/r2/api/workers/workers-api-usage/) |
| Create, resume, complete, or abort multipart uploads | [Multipart Worker and client example](https://developers.cloudflare.com/r2/api/workers/workers-multipart-usage/) |
| Check supported S3 operations and headers | [S3 compatibility](https://developers.cloudflare.com/r2/api/s3/api/) |
| Configure an S3 JavaScript client | [AWS SDK for JavaScript v3](https://developers.cloudflare.com/r2/examples/aws/aws-sdk-js-v3/) |
| Sign temporary upload or download access | [Presigned URLs](https://developers.cloudflare.com/r2/api/s3/presigned-urls/) |
| Encrypt with customer-provided keys | [SSE-C usage](https://developers.cloudflare.com/r2/examples/ssec/) |
| Manage buckets and objects from the command line | [Wrangler R2 commands](https://developers.cloudflare.com/r2/reference/wrangler-commands/) |

Use generated project types rather than maintaining local copies of R2 interfaces; see [Workers TypeScript guidance](https://developers.cloudflare.com/workers/languages/typescript/). For pagination and conditional response handling, read [gotchas.md](./gotchas.md) alongside the API reference.

# R2 Configuration

Fetch the task's documentation before editing Wrangler configuration or bucket settings.

| Task | Current documentation |
|------|-----------------------|
| Create a bucket and bind it to a Worker | [Workers API setup](https://developers.cloudflare.com/r2/get-started/workers-api/) |
| Choose local simulation or a remote bucket during development | [Supported bindings per development mode](https://developers.cloudflare.com/workers/local-development/bindings-per-env/) and [local development](https://developers.cloudflare.com/workers/local-development/) |
| Create S3 credentials and scope permissions | [R2 authentication](https://developers.cloudflare.com/r2/api/tokens/) |
| Set the S3 endpoint and SDK region | [AWS SDK for JavaScript v3](https://developers.cloudflare.com/r2/examples/aws/aws-sdk-js-v3/) |
| Choose placement hints or a jurisdiction | [Data location](https://developers.cloudflare.com/r2/reference/data-location/) |
| Configure browser origins, methods, and headers | [CORS](https://developers.cloudflare.com/r2/buckets/cors/) |
| Set expiration, storage transitions, or incomplete-upload cleanup | [Object lifecycles](https://developers.cloudflare.com/r2/buckets/object-lifecycles/) |
| Choose or change storage classes | [Storage classes](https://developers.cloudflare.com/r2/buckets/storage-classes/) and [pricing](https://developers.cloudflare.com/r2/pricing/) |
| Send object events to a queue | [Event notifications](https://developers.cloudflare.com/r2/buckets/event-notifications/) |
| Configure public access or a custom domain | [Public buckets](https://developers.cloudflare.com/r2/buckets/public-buckets/) |
| Manage bucket settings with Wrangler | [R2 commands](https://developers.cloudflare.com/r2/reference/wrangler-commands/) |

Choose the development bucket deliberately: a remote binding accesses real data. Scope S3 credentials to the required buckets and operations; Workers bindings use their own access mechanism. Review lifecycle prefixes and retention needs before applying deletion rules, and evaluate retrieval and minimum-storage charges before choosing a storage class.

# R2 Gotchas & Troubleshooting

Use the current references to diagnose the actual response or error instead of copying a workaround.

| Symptom or decision | What to check |
|---------------------|---------------|
| Listing stops early or escapes the intended prefix | Follow `truncated` and the returned cursor, retaining the original prefix, delimiter, and metadata options on subsequent requests. See the listing section of the [Workers API reference](https://developers.cloudflare.com/r2/api/workers/workers-api-reference/). |
| Conditional read has no body, or conditional write returns null | Distinguish a missing object from a failed condition; choose the HTTP response for the actual request condition. See conditional operations in the [Workers API reference](https://developers.cloudflare.com/r2/api/workers/workers-api-reference/). |
| ETag, metadata, checksum, or stream upload behaves unexpectedly | Check supported values and return types in the [Workers API reference](https://developers.cloudflare.com/r2/api/workers/workers-api-reference/), the [Worker upload example](https://developers.cloudflare.com/r2/api/workers/workers-api-usage/), and [Workers streams](https://developers.cloudflare.com/workers/runtime-apis/streams/). |
| Multipart upload fails or cannot be resumed | Check part constraints and handle an upload that has already completed or aborted: [multipart guide](https://developers.cloudflare.com/r2/api/workers/workers-multipart-usage/) and [API reference](https://developers.cloudflare.com/r2/api/workers/workers-api-reference/). |
| S3 authentication or signed URL fails | Verify credentials, endpoint, region, operation, signed headers, and expiry using [SDK setup](https://developers.cloudflare.com/r2/examples/aws/aws-sdk-js-v3/), [authentication](https://developers.cloudflare.com/r2/api/tokens/), and [presigned URLs](https://developers.cloudflare.com/r2/api/s3/presigned-urls/). |
| Browser fails but an HTTP client succeeds | Check [CORS](https://developers.cloudflare.com/r2/buckets/cors/) and [troubleshooting](https://developers.cloudflare.com/r2/platform/troubleshooting/). |
| Local and deployed data or behavior differ | Check [local development](https://developers.cloudflare.com/workers/local-development/), [supported bindings](https://developers.cloudflare.com/workers/local-development/bindings-per-env/), and the local persistence options in [Wrangler R2 commands](https://developers.cloudflare.com/r2/reference/wrangler-commands/). |
| Reads serve old or missing content after an update | Check the [consistency model and cache interactions](https://developers.cloudflare.com/r2/reference/consistency/). |
| Upload size, metadata size, storage cost, or lifecycle behavior is unexpected | Fetch [limits](https://developers.cloudflare.com/r2/platform/limits/), [pricing](https://developers.cloudflare.com/r2/pricing/), [storage classes](https://developers.cloudflare.com/r2/buckets/storage-classes/), and [object lifecycles](https://developers.cloudflare.com/r2/buckets/object-lifecycles/). |

For other failures, start with [R2 troubleshooting](https://developers.cloudflare.com/r2/platform/troubleshooting/) and [error codes](https://developers.cloudflare.com/r2/api/error-codes/).

# R2 Patterns & Best Practices

Choose the access and delivery model, then fetch the implementation guide.

| Task | Current documentation |
|------|-----------------------|
| Stream object downloads or accept uploads through a Worker | [Use R2 from Workers](https://developers.cloudflare.com/r2/api/workers/workers-api-usage/) |
| Add conditional reads/writes, range handling, checksums, or batch deletion | [Workers API reference](https://developers.cloudflare.com/r2/api/workers/workers-api-reference/) |
| Upload large files with multipart state tracked by the client | [Multipart Worker and client example](https://developers.cloudflare.com/r2/api/workers/workers-multipart-usage/) |
| Upload directly from a browser or share a temporary download | [Presigned URLs](https://developers.cloudflare.com/r2/api/s3/presigned-urls/) and [CORS](https://developers.cloudflare.com/r2/buckets/cors/) |
| Cache responses served by a Worker | [Cache API example](https://developers.cloudflare.com/r2/examples/cache-api/) |
| Deliver public objects through a custom domain or evaluate r2.dev | [Public buckets](https://developers.cloudflare.com/r2/buckets/public-buckets/) |
| Process object changes asynchronously | [Event notifications](https://developers.cloudflare.com/r2/buckets/event-notifications/) and [Queues](../queues/) |
| Expire objects or transition storage classes | [Object lifecycles](https://developers.cloudflare.com/r2/buckets/object-lifecycles/) and [storage classes](https://developers.cloudflare.com/r2/buckets/storage-classes/) |

Authorize the caller for the selected object key and operation before exposing a Worker endpoint or issuing a presigned URL. A key-format check alone does not establish access rights. Set the intended expiry and signed request constraints for temporary access; configure browser CORS separately.

Keep private responses out of shared public caches. Choose cache keys and invalidation around the application's access model, and check [R2 consistency and caching behavior](https://developers.cloudflare.com/r2/reference/consistency/) when objects can change. For multipart uploads, plan for failed parts, completion, and cleanup using the linked guide.

# Cloudflare R2 Object Storage

Use R2 for objects such as uploads, media, backups, and static assets. Fetch the linked documentation before implementing; API signatures, configuration, limits, and pricing belong in the current docs.

## Choose an access path

- Use a Workers binding for object access inside a Worker: [Workers API setup](https://developers.cloudflare.com/r2/get-started/workers-api/).
- Use the S3-compatible API for existing S3 clients or direct client access through presigned URLs: [S3 setup](https://developers.cloudflare.com/r2/get-started/s3/). Check supported operations rather than assuming full S3 parity.
- Decide whether objects need application authorization, temporary access, or public delivery before exposing the bucket. See [patterns.md](./patterns.md).

## Find the task

| Task | Reference |
|------|-----------|
| Bindings, credentials, local development, bucket settings | [configuration.md](./configuration.md) |
| Object operations, metadata, conditions, multipart, CLI | [api.md](./api.md) |
| Uploads, streaming, caching, public delivery, event processing | [patterns.md](./patterns.md) |
| Pagination, conditional responses, failed uploads, limits | [gotchas.md](./gotchas.md) |

For other topics, discover pages through the [R2 documentation index](https://developers.cloudflare.com/r2/llms.txt). Check [pricing](https://developers.cloudflare.com/r2/pricing/) before estimating costs.

## See also

- [Workers](https://developers.cloudflare.com/workers/) for request handling.
- [KV](../kv/) or [D1](../d1/) for application metadata associated with objects.
- [Queues](../queues/) for asynchronous processing of object events.

# R2 Data Catalog API Selection

Use the Iceberg REST catalog through an engine for table reads and writes; use the Cloudflare control-plane API for catalog administration. Copy the catalog connection values from the actual environment as described in [configuration](configuration.md).

| Task | Documentation |
|------|---------------|
| Enable or disable catalogs; inspect status, credentials, namespaces, tables, and maintenance configuration | [R2 Data Catalog control-plane API](https://developers.cloudflare.com/api/resources/r2_data_catalog/) — select the affected operation for its schema, pagination, and namespace encoding |
| Connect and create tables through Python | [PyIceberg configuration](https://developers.cloudflare.com/r2-data-catalog/config-examples/pyiceberg/) |
| Connect, create, write, and query through Spark | [PySpark configuration](https://developers.cloudflare.com/r2-data-catalog/config-examples/spark-python/) |
| Plan automatic compaction and snapshot expiration | [Table maintenance](https://developers.cloudflare.com/r2-data-catalog/table-maintenance/) |
| Delete rows, tables, or associated files | [Deleting data](https://developers.cloudflare.com/r2-data-catalog/deleting-data/) |

For engine-specific operations beyond these Cloudflare examples, follow the upstream engine documentation linked from the relevant configuration guide and check the installed version. Do not infer engine method signatures from the control-plane API.

## Get Table (repository-specific metadata introspection note)

This existing repository note is retained because the published control-plane API reference does not document this operation or its snapshot-pruning response. Verify availability and response behavior against the target service or authoritative implementation before relying on it; it is not a documented API guarantee. Do not substitute the documented list-tables response for this metadata response.

`GET /namespaces/{ns}/tables/{table}` returns schema, partition spec, sort order, and snapshot info — like Iceberg "load table" but on the control plane, with snapshots pruned to the most recent 10.

```bash
curl -s "https://api.cloudflare.com/client/v4/accounts/$ACCOUNT_ID/r2-catalog/$BUCKET/namespaces/live/tables/earthquakes" \
  -H "Authorization: Bearer $API_TOKEN"
```

```json
{"result": {
  "identifier": {"namespace": ["live"], "name": "earthquakes"},
  "table_uuid": "019edccf-3ac8-73e3-...",
  "metadata_location": "s3://live-data/__r2_data_catalog/.../metadata/01225-....metadata.json",
  "total_snapshots": 1225,
  "returned_snapshots": 10,
  "metadata": { /* standard Iceberg TableMetadata: schemas, partition-specs, sort-orders,
                   properties, current-snapshot-id, snapshots (≤10), snapshot-log, refs */ }
}, "success": true}
```

| Field | Description |
|-------|-------------|
| `identifier` | `{namespace: [...], name}` |
| `table_uuid` | Iceberg table UUID |
| `metadata_location` | R2 path to current metadata file |
| `total_snapshots` | Total before pruning |
| `returned_snapshots` | Count in `metadata.snapshots` (max 10) |
| `metadata` | Standard [Iceberg TableMetadata](https://iceberg.apache.org/spec/#table-metadata-fields), arrays pruned to 10 |

See [patterns](patterns.md) for engine selection and [troubleshooting](gotchas.md) for diagnosis.

# R2 Data Catalog Configuration

Inspect the existing bucket, catalog, engine versions, and credential configuration before making changes. Use the project's installed tools and preserve its environment-variable or secret-management conventions.

| Task | Documentation |
|------|---------------|
| Enable a catalog and obtain connection details | [Enable R2 Data Catalog](https://developers.cloudflare.com/r2-data-catalog/manage-catalogs/#enable-r2-data-catalog-on-a-bucket) |
| Select credentials for readers, writers, or maintenance | [Authenticate your Iceberg engine](https://developers.cloudflare.com/r2-data-catalog/manage-catalogs/#authenticate-your-iceberg-engine) |
| Configure compaction and its service credential | [Enable compaction](https://developers.cloudflare.com/r2-data-catalog/manage-catalogs/#enable-compaction) |
| Configure snapshot retention | [Enable snapshot expiration](https://developers.cloudflare.com/r2-data-catalog/manage-catalogs/#enable-snapshot-expiration) |
| Choose file sizes, retention policy, and maintenance scope | [Table maintenance](https://developers.cloudflare.com/r2-data-catalog/table-maintenance/) |
| Connect a Python client | [PyIceberg](https://developers.cloudflare.com/r2-data-catalog/config-examples/pyiceberg/) |
| Connect Spark | [PySpark](https://developers.cloudflare.com/r2-data-catalog/config-examples/spark-python/) |
| Connect another query engine | [Engine configuration guides](https://developers.cloudflare.com/r2-data-catalog/config-examples/) |
| Disable catalog access | [Disable R2 Data Catalog](https://developers.cloudflare.com/r2-data-catalog/manage-catalogs/#disable-r2-data-catalog-on-a-bucket) |

Copy the Catalog URI and Warehouse name exactly from the catalog detail page or Wrangler enable output. Scope both catalog and storage permissions to the operations the client needs; readers do not need a blanket write-enabled token. Treat maintenance credentials separately from reader credentials. Verify connectivity with a read operation before attempting writes, then check catalog and credential status through the [control-plane API](api.md).

# R2 Data Catalog Troubleshooting

Identify whether the failure occurs in catalog administration, engine metadata access, or underlying object access before changing permissions or client settings.

| Check | Documentation |
|-------|---------------|
| Catalog enablement, Catalog URI, or Warehouse mismatch | [Manage catalogs](https://developers.cloudflare.com/r2-data-catalog/manage-catalogs/) |
| Reader/writer token scope or file-access denial | [Engine authentication](https://developers.cloudflare.com/r2-data-catalog/manage-catalogs/#authenticate-your-iceberg-engine) — inspect both catalog and storage permissions |
| Missing maintenance credentials or wrong table/catalog configuration | [Control-plane API](https://developers.cloudflare.com/api/resources/r2_data_catalog/) and [enable compaction](https://developers.cloudflare.com/r2-data-catalog/manage-catalogs/#enable-compaction) |
| Compaction backlog, retention, or orphaned files | [Table maintenance](https://developers.cloudflare.com/r2-data-catalog/table-maintenance/) |
| PyIceberg connection or table creation | [PyIceberg configuration](https://developers.cloudflare.com/r2-data-catalog/config-examples/pyiceberg/) |
| Spark dependency, credential-vending, or signing configuration | [PySpark configuration](https://developers.cloudflare.com/r2-data-catalog/config-examples/spark-python/) |
| Deleted data is still present | [Deleting data](https://developers.cloudflare.com/r2-data-catalog/deleting-data/) |
| Catalog request or maintenance-job diagnosis | [Metrics and analytics](https://developers.cloudflare.com/r2-data-catalog/observability/metrics/) |

Compare the client's configured URI and warehouse with the actual catalog values. Test a read operation first; do not grant write access merely to resolve a reader's failure. For schema or concurrency errors, inspect the installed engine's behavior and current table metadata before retrying. The [get-table note](api.md#get-table-repository-specific-metadata-introspection-note) is not a substitute for verifying the service's response contract.

See [configuration](configuration.md) and [patterns](patterns.md) for implementation choices.

# R2 Data Catalog Patterns

Choose the engine based on the project's existing runtime and workload, then retrieve its current connection example.

| Need | Starting point |
|------|----------------|
| Python catalog operations and ingestion without a Spark deployment | [PyIceberg](https://developers.cloudflare.com/r2-data-catalog/config-examples/pyiceberg/) |
| Existing Spark ETL and distributed table processing | [PySpark](https://developers.cloudflare.com/r2-data-catalog/config-examples/spark-python/) |
| Connect an existing SQL engine | [Engine configuration guides](https://developers.cloudflare.com/r2-data-catalog/config-examples/) |
| Query through Cloudflare's serverless SQL service | [R2 SQL](../r2-sql/) |
| Stream events into tables | [Pipelines patterns](../pipelines/patterns.md) |

Use the discovered Catalog URI and Warehouse name from [configuration](configuration.md). Match dependencies to the installed engine and the current guide instead of adopting a universal pinned Spark/Iceberg combination.

Plan ingestion, query, and maintenance responsibilities together. Prefer [automatic table maintenance](https://developers.cloudflare.com/r2-data-catalog/table-maintenance/) when it meets the workload; align retention with time-travel needs before enabling expiration. For engine-specific partitioning, schema evolution, or manual procedures, consult that engine's linked upstream documentation and verify behavior on representative data.

When multiple writers share a table, design recovery around the actual failed operation and the engine's commit semantics. Reproduce conflicts and ensure retries do not duplicate application work. See [API selection](api.md) and [troubleshooting](gotchas.md).

# R2 Data Catalog

Use R2 Data Catalog for Iceberg analytics and data pipelines on object storage. For transactional application queries, consider a database; for unstructured objects, use [R2](../r2/).

Distinguish the Iceberg REST catalog used by query engines from Cloudflare's control-plane API for catalog administration. Start with the workflow you need:

| Task | Reference |
|------|-----------|
| Enable a catalog, discover connection values, and choose credentials | [Configuration](configuration.md) |
| Select administration or engine APIs | [API selection](api.md) |
| Choose a Python, Spark, or SQL workflow | [Patterns](patterns.md) |
| Diagnose authentication, maintenance, or client problems | [Troubleshooting](gotchas.md) |

Copy the actual **Catalog URI** and **Warehouse name** from the catalog detail page or Wrangler's enable output. Pass both to the selected engine; do not reconstruct them from an assumed bucket naming convention. Retrieve [Manage catalogs](https://developers.cloudflare.com/r2-data-catalog/manage-catalogs/) before setup or permission changes.

Related workflows: [Pipelines](../pipelines/) for ingest and [R2 SQL](../r2-sql/) for querying tables.

# R2 SQL API Reference

Read-only SQL over Iceberg (Apache DataFusion). Query templates only. For the authoritative list of supported syntax, functions, data types, and limitations, pull the SQL reference (`sql-reference/`, `.../aggregate-functions/`, `.../scalar-functions/`, `.../complex-types/`) and `reference/limitations-best-practices/`.

## Query Endpoint

```
POST https://api.sql.cloudflarestorage.com/api/v1/accounts/{ACCOUNT_ID}/r2-sql/query/{BUCKET}
Authorization: Bearer <token>
Content-Type: application/json
Body: {"query": "<SQL>"}
```

CLI: `npx wrangler r2 sql query "{WAREHOUSE}" "<SQL>"` (with `WRANGLER_R2_SQL_AUTH_TOKEN`).

## Response Format

```json
{
  "result": {
    "request_id": "dqe-prod-01...",
    "schema": [{"name": "cnt", "descriptor": {"type": {"name": "int64"}, "nullable": false}}],
    "rows": [{"category": "Electronics", "cnt": 12345}],
    "metrics": {"r2_requests_count": 5, "files_scanned": 29, "bytes_scanned": 12345678, "cache_hits": 0}
  },
  "success": true, "errors": []
}
```

Error: `{"result": null, "success": false, "errors": [{"code": 40003, "message": "..."}]}`. `bytes_scanned` ≈ billable data.

## Query Structure

```sql
SELECT [DISTINCT] columns | expressions | aggregations
FROM namespace.table [alias]
[ [INNER|LEFT|RIGHT|FULL OUTER|CROSS] JOIN namespace.table2 alias2 ON ... ]
[WHERE ...] [GROUP BY ...] [HAVING ...]
[QUALIFY window_predicate]
[ORDER BY expr [ASC|DESC]]
[LIMIT n]                          -- default 500, max 10,000
```

## Schema Discovery

```sql
SHOW DATABASES;            -- list namespaces (aliases: SHOW NAMESPACES / SHOW SCHEMAS)
SHOW TABLES IN namespace;
DESCRIBE namespace.table;  -- columns, types, partition keys
EXPLAIN [FORMAT JSON] SELECT ...;   -- execution plan (free; no data scanned)
```

## JOINs / Subqueries / CTEs / Set Ops

```sql
-- JOINs: all types + multi-way
SELECT z.domain, COUNT(*) AS cnt
FROM ns.zones z
INNER JOIN ns.http_requests h ON z.zone_id = h.zone_id
LEFT  JOIN ns.firewall_events f ON z.zone_id = f.zone_id
GROUP BY z.domain ORDER BY cnt DESC LIMIT 20;

-- Subqueries: IN / EXISTS / scalar / derived
SELECT * FROM ns.t1 WHERE id IN (SELECT id FROM ns.t2 WHERE x > 0);
SELECT col, (SELECT COUNT(*) FROM ns.t2 s WHERE s.id = t.id) AS cnt FROM ns.t1 t;

-- Multi-table CTE with JOIN
WITH top AS (SELECT zone_id, COUNT(*) AS req FROM ns.http_requests GROUP BY zone_id ORDER BY req DESC LIMIT 50)
SELECT t.zone_id, t.req FROM top t LEFT JOIN ns.zones z ON t.zone_id = z.zone_id;

-- Set ops: UNION / UNION ALL / INTERSECT / EXCEPT
SELECT zone_id FROM ns.firewall_events WHERE action = 'block'
UNION SELECT zone_id FROM ns.http_requests WHERE risk_score > 0.8;
```

## Window Functions

Use inline `OVER (...)`. See the SQL reference for the full list of supported window functions and frame syntax.

```sql
SELECT event_id,
       ROW_NUMBER() OVER (PARTITION BY mag_type ORDER BY magnitude DESC) AS rn,
       LAG(magnitude, 2, 0.0) OVER (ORDER BY occurred_at) AS prev2,   -- offset + default
       NTH_VALUE(magnitude, 2) OVER (ORDER BY magnitude DESC) AS n2,
       SUM(magnitude) OVER (ORDER BY occurred_at) AS running,
       AVG(magnitude) OVER (ORDER BY magnitude ROWS BETWEEN 2 PRECEDING AND CURRENT ROW) AS moving_avg
FROM ns.earthquakes;

-- QUALIFY: filter on a window result (top row per partition)
SELECT event_id, mag_type, magnitude FROM ns.earthquakes
QUALIFY ROW_NUMBER() OVER (PARTITION BY mag_type ORDER BY magnitude DESC) = 1;
```

## Functions

Aggregate, scalar, JSON, and array/map function catalogs are in the docs — pull `sql-reference/aggregate-functions/` and `.../scalar-functions/`. JSON functions accept variadic paths, e.g. `json_get_int(doc, 'user', 'profile', 'level')`.

## Data Types

`integer`, `float`, `string` (single quotes), `boolean`, `timestamp` (RFC3339 **with timezone**), `date` (ISO 8601), `struct`, `array` (1-indexed), `map`. No implicit conversions — quote strings, include timezone on timestamps, don't quote integers. Full type docs: `sql-reference/`.

```sql
WHERE status = 200 AND method = 'GET'              -- not '200', not GET
  AND ts >= '2026-01-01T00:00:00Z'                 -- not '2026-01-01'
```

## Complex Types (quick examples; full ref in docs)

```sql
SELECT pricing['price'] AS price, get_field(pricing, 'discount') AS disc FROM ns.t;  -- struct
SELECT tags[1] AS first_tag, array_length(tags) AS n FROM ns.t;                       -- array (1-indexed)
SELECT map_keys(meta), map_extract(meta, 'source') FROM ns.t;                         -- map
```

## Errors

Failed queries return `{"success": false, "errors": [{"code": ..., "message": ...}]}`. For error codes and troubleshooting, see `https://developers.cloudflare.com/r2-sql/troubleshooting/`.

## See Also

- [patterns.md](patterns.md) — query examples · [gotchas.md](gotchas.md) — limits & workarounds · [configuration.md](configuration.md)

# R2 SQL Configuration

Auth and setup. For the current permission matrix and wrangler flags, pull `https://developers.cloudflare.com/r2-sql/reference/wrangler-commands/` and the R2 Data Catalog manage-catalogs doc.

## Prerequisites

- R2 bucket with Data Catalog enabled ([r2-data-catalog/configuration.md](../r2-data-catalog/configuration.md))
- R2 API token: **R2 Storage Admin Read & Write** (includes R2 SQL Read), or add **R2 SQL Read** explicitly
- Wrangler CLI (for CLI queries)

> Open-beta limitation: R2 Storage **Admin Read & Write is required even for read-only R2 SQL queries**.

## Enable Catalog + Get Warehouse

```bash
npx wrangler r2 bucket catalog enable my-bucket
```

You query by **warehouse** name (`{ACCOUNT_ID}_{BUCKET}`), shown in the output alongside the Catalog URI.

## Configure Auth

### Wrangler CLI

```bash
export WRANGLER_R2_SQL_AUTH_TOKEN=<your-token>
# or a .env file in the project dir (auto-loaded): WRANGLER_R2_SQL_AUTH_TOKEN=<your-token>
```

> Wrangler does **not** use the `wrangler login` OAuth session for R2 SQL — the env var is required.

### REST API

```bash
curl -X POST \
  "https://api.sql.cloudflarestorage.com/api/v1/accounts/$ACCOUNT_ID/r2-sql/query/$BUCKET" \
  -H "Authorization: Bearer $TOKEN" -H "Content-Type: application/json" \
  -d '{"query": "SELECT * FROM default.my_table LIMIT 10"}'
```

## Verify Setup

```bash
npx wrangler r2 sql query "${ACCOUNT_ID}_my-bucket" "SHOW DATABASES"
npx wrangler r2 sql query "${ACCOUNT_ID}_my-bucket" "SHOW TABLES IN default"
```

## See Also

- [api.md](api.md) — SQL syntax · [patterns.md](patterns.md) — query examples · [gotchas.md](gotchas.md) — troubleshooting

# R2 SQL Gotchas

Operational pitfalls. For the authoritative list of supported features, unsupported features, and recommended workarounds, pull `https://developers.cloudflare.com/r2-sql/reference/limitations-best-practices/` and `https://developers.cloudflare.com/r2-sql/troubleshooting/`.

## Access

- **No Workers binding.** There is no `env.R2_SQL`. Query the REST endpoint via `fetch()` from a Worker ([patterns.md](patterns.md#dashboard-worker)), or use D1 / an external DB for OLTP.
- Wrangler needs `WRANGLER_R2_SQL_AUTH_TOKEN` — it does **not** reuse the `wrangler login` OAuth session.
- Open beta: R2 Storage **Admin Read & Write is required even for read-only** queries.

## Type Safety

```sql
-- ❌ wrong                          -- ✅ right
WHERE status = '200'                 WHERE status = 200
WHERE ts > '2026-01-01'              WHERE ts > '2026-01-01T00:00:00Z'   -- need time + tz
WHERE method = GET                   WHERE method = 'GET'
```

No implicit conversions. Timestamps must be RFC3339 with timezone; dates ISO 8601.

## Performance

- **File count dominates latency** — enable automatic compaction.
- **Partition-filter + narrow time windows + always `LIMIT`.**
- **Multi-way JOINs on large tables** can exceed resource limits — filter heavily, join through dimension tables.
- Per-query `metrics` (`files_scanned`, `bytes_scanned`, `cache_hits`) are the primary observability signal; `bytes_scanned` ≈ billable data. For LIMIT bounds, pagination, and other guidance, see the limitations-best-practices doc.

## Debug Checklist

1. `wrangler r2 bucket catalog enable <bucket>` — catalog on?
2. `echo $WRANGLER_R2_SQL_AUTH_TOKEN` — token set?
3. `SHOW DATABASES` → `SHOW TABLES IN ns` → `DESCRIBE ns.table`
4. `SELECT COUNT(*) FROM ns.table` — data present?
5. Add filters incrementally; read `metrics` to tune.

## See Also

- [api.md](api.md) · [patterns.md](patterns.md) · [configuration.md](configuration.md)

# R2 SQL Patterns

Code templates for CLI, REST, and Worker access. For performance/partitioning best practices, pull `https://developers.cloudflare.com/r2-sql/reference/limitations-best-practices/`.

## Wrangler CLI

```bash
export WRANGLER_R2_SQL_AUTH_TOKEN=$API_TOKEN

npx wrangler r2 sql query "${ACCOUNT_ID}_my-bucket" "
  SELECT category, COUNT(*) AS cnt, round(AVG(amount), 2) AS avg_amount
  FROM analytics.events
  WHERE __ingest_ts >= '2026-01-01T00:00:00Z'
  GROUP BY category ORDER BY cnt DESC LIMIT 100"
```

## REST API (Python)

```python
import requests

API = f"https://api.sql.cloudflarestorage.com/api/v1/accounts/{ACCOUNT_ID}/r2-sql/query/{BUCKET}"
HEADERS = {"Authorization": f"Bearer {TOKEN}", "Content-Type": "application/json"}

def r2sql(query):
    body = requests.post(API, headers=HEADERS, json={"query": query}, timeout=180).json()
    if body["success"]:
        return body["result"]["rows"], body["result"]["metrics"]
    raise RuntimeError(body["errors"])

rows, metrics = r2sql("SELECT category, COUNT(*) AS cnt FROM analytics.events GROUP BY category LIMIT 10")
```

## REST API (curl)

```bash
curl -X POST \
  "https://api.sql.cloudflarestorage.com/api/v1/accounts/$ACCOUNT_ID/r2-sql/query/$BUCKET" \
  -H "Authorization: Bearer $TOKEN" -H "Content-Type: application/json" \
  -d '{"query": "SELECT COUNT(*) AS total FROM analytics.events"}'
```

## Dashboard Worker

No R2 SQL binding exists — query the REST endpoint via `fetch()`.

```typescript
interface Env { ACCOUNT_ID: string; BUCKET: string; R2_SQL_TOKEN: string; }

async function queryR2SQL(env: Env, query: string) {
  const url = `https://api.sql.cloudflarestorage.com/api/v1/accounts/${env.ACCOUNT_ID}/r2-sql/query/${env.BUCKET}`;
  const resp = await fetch(url, {
    method: "POST",
    headers: { Authorization: `Bearer ${env.R2_SQL_TOKEN}`, "Content-Type": "application/json" },
    body: JSON.stringify({ query }),
  });
  if (!resp.ok) throw new Error(`R2 SQL ${resp.status}: ${await resp.text()}`);
  return (await resp.json() as any).result;
}

export default {
  async fetch(req: Request, env: Env): Promise<Response> {
    if (new URL(req.url).pathname === "/api/analytics") {
      const result = await queryR2SQL(env, `
        SELECT category, COUNT(*) AS cnt FROM analytics.events
        GROUP BY category ORDER BY cnt DESC LIMIT 10`);
      return Response.json(result.rows);
    }
    return new Response("Not found", { status: 404 });
  },
};
```

```bash
npx wrangler secret put R2_SQL_TOKEN
```

## Example Queries

```sql
-- Error rate by endpoint
SELECT path, COUNT(*) AS total, SUM(CASE WHEN status >= 400 THEN 1 ELSE 0 END) AS errors
FROM logs.http_requests WHERE __ingest_ts >= '2026-01-01T00:00:00Z'
GROUP BY path ORDER BY errors DESC LIMIT 20;

-- Top-3 slowest requests per method (window + QUALIFY)
SELECT method, path, response_time_ms FROM logs.http_requests
QUALIFY ROW_NUMBER() OVER (PARTITION BY method ORDER BY response_time_ms DESC) <= 3;

-- Cross-table analytics with approx distinct
SELECT z.domain, COUNT(*) AS requests, approx_distinct(h.client_ip) AS uniques
FROM ns.zones z INNER JOIN ns.http_requests h ON z.zone_id = h.zone_id
WHERE h.__ingest_ts >= '2026-06-01T00:00:00Z'
GROUP BY z.domain ORDER BY requests DESC LIMIT 25;
```

## Cursor-Based Pagination

Paginate on a sortable (ideally partition) column rather than `OFFSET`:

```sql
SELECT * FROM logs.requests ORDER BY __ingest_ts DESC LIMIT 500;                       -- page 1
SELECT * FROM logs.requests WHERE __ingest_ts < '<last_ts>' ORDER BY __ingest_ts DESC LIMIT 500;  -- page 2
```

## Performance (essentials)

- **Always `LIMIT`** (early termination); **filter on partition keys first** (`__ingest_ts` range), then add predicates.
- **Narrow time ranges**; **compact tables** (file count dominates latency — enable automatic compaction in [r2-data-catalog](../r2-data-catalog/configuration.md)).
- Read response `metrics` (`files_scanned`, `bytes_scanned`) to tune. Full guidance: limitations-best-practices doc.

## Pipelines → R2 SQL

After `npx wrangler pipelines setup` (Data Catalog destination), wait for first flush (3–7 min), then query the table. See [pipelines/patterns.md](../pipelines/patterns.md).

## See Also

- [api.md](api.md) · [gotchas.md](gotchas.md) · [r2-data-catalog/patterns.md](../r2-data-catalog/patterns.md)

# Cloudflare R2 SQL

Serverless, distributed, **read-only** query engine (Apache DataFusion) for Apache Iceberg tables in R2 Data Catalog.

## Documentation

For full function lists, data types, and pricing, **retrieve the live docs** — use the Cloudflare MCP `docs` tool if available, otherwise `webfetch`.

| Topic | URL |
|-------|-----|
| Overview / get started | `https://developers.cloudflare.com/r2-sql/get-started/` |
| Query data | `https://developers.cloudflare.com/r2-sql/query-data/` |
| SQL reference | `https://developers.cloudflare.com/r2-sql/sql-reference/` |
| Aggregate functions | `https://developers.cloudflare.com/r2-sql/sql-reference/aggregate-functions/` |
| Scalar functions | `https://developers.cloudflare.com/r2-sql/sql-reference/scalar-functions/` |
| Complex types | `https://developers.cloudflare.com/r2-sql/sql-reference/complex-types/` |
| Limitations & best practices | `https://developers.cloudflare.com/r2-sql/reference/limitations-best-practices/` |
| Wrangler commands | `https://developers.cloudflare.com/r2-sql/reference/wrangler-commands/` |
| Pricing | `https://developers.cloudflare.com/r2-sql/platform/pricing/` |

## Connection Values

| Value | Format |
|-------|--------|
| REST endpoint | `https://api.sql.cloudflarestorage.com/api/v1/accounts/{ACCOUNT_ID}/r2-sql/query/{BUCKET}` |
| Wrangler | `npx wrangler r2 sql query "{WAREHOUSE}" "<SQL>"` with `WRANGLER_R2_SQL_AUTH_TOKEN` set |
| Warehouse | `{ACCOUNT_ID}_{BUCKET}` |

> The REST endpoint is `api.sql.cloudflarestorage.com` — **not** `api.cloudflare.com/.../r2/sql`.

## Quick Start

```bash
npx wrangler r2 bucket catalog enable my-bucket           # 1. enable catalog
export WRANGLER_R2_SQL_AUTH_TOKEN=<r2-token>              # 2. auth (Admin R&W + R2 SQL Read)
npx wrangler r2 sql query "$ACCOUNT_ID"_my-bucket \
  "SELECT * FROM default.my_table LIMIT 10"                # 3. query
```

## SQL Surface

R2 SQL is read-only and supports a broad analytical SQL surface (SELECT, JOINs, subqueries, CTEs, set operations, window functions, and aggregate/scalar/JSON functions over complex types). For the authoritative, current list of supported syntax, functions, and limitations, see the SQL reference and limitations docs linked above. [api.md](api.md) has query templates.

## When to Use

**Use for:** SQL analytics over Iceberg (logs, BI, fraud, ad-hoc), multi-cloud queries without egress, dashboards (query from a Worker via HTTP).

**Don't use for:** writes (use PySpark/PyIceberg) or real-time OLTP (<100 ms).

## No Workers Binding

There is no `env.R2_SQL` binding. Query from a Worker via `fetch()` to the REST endpoint with the token as a secret (see [patterns.md](patterns.md#dashboard-worker)).

## Reading Order

1. [configuration.md](configuration.md) — enable catalog, tokens, env setup
2. [api.md](api.md) — SQL syntax templates, JOIN/window examples, response format, data types
3. [patterns.md](patterns.md) — CLI/REST/Worker queries, use cases, pagination, performance
4. [gotchas.md](gotchas.md) — what works vs. not, performance, troubleshooting

## See Also

- [r2-data-catalog](../r2-data-catalog/) — PyIceberg/PySpark, table management
- [pipelines](../pipelines/) — streaming ingest into queryable tables

# API Reference

## Authentication

```bash
curl -X POST 'https://rtc.live/v1/apps/${CALLS_APP_ID}/sessions/new' \
  -H "Authorization: Bearer ${CALLS_APP_SECRET}"
```

## Core Concepts

**Sessions:** PeerConnection to Cloudflare edge  
**Tracks:** Media/data channels (audio/video/datachannel)  
**No rooms:** Build presence via track sharing

## Client Libraries

**PartyTracks (Recommended):** Observable-based client library for production use. Handles device changes, network switches, ICE restarts automatically. Push/pull API with React hooks. See patterns.md for full examples.

```bash
npm install partytracks @cloudflare/calls
```

**Raw API:** Direct HTTP + WebRTC for custom requirements (documented below).

## Endpoints

### Create Session
```http
POST /v1/apps/{appId}/sessions/new
→ {sessionId, sessionDescription}
```

### Add Track (Publish)
```http
POST /v1/apps/{appId}/sessions/{sessionId}/tracks/new
Body: {
  sessionDescription: {sdp, type: "offer"},
  tracks: [{location: "local", trackName: "my-video"}]
}
→ {sessionDescription, tracks: [{trackName}]}
```

### Add Track (Subscribe)
```http
POST /v1/apps/{appId}/sessions/{sessionId}/tracks/new
Body: {
  tracks: [{
    location: "remote",
    trackName: "remote-track-id",
    sessionId: "other-session-id"
  }]
}
→ {sessionDescription} (server offer)
```

### Renegotiate
```http
PUT /v1/apps/{appId}/sessions/{sessionId}/renegotiate
Body: {sessionDescription: {sdp, type: "answer"}}
```

### Close Tracks
```http
PUT /v1/apps/{appId}/sessions/{sessionId}/tracks/close
Body: {tracks: [{trackName}]}
→ {requiresImmediateRenegotiation: boolean}
```

### Get Session
```http
GET /v1/apps/{appId}/sessions/{sessionId}
→ {sessionId, tracks: TrackMetadata[]}
```

## TypeScript Types

```typescript
interface TrackMetadata {
  trackName: string;
  location: "local" | "remote";
  sessionId?: string; // For remote tracks
  mid?: string; // WebRTC mid
}
```

## WebRTC Flow

```typescript
// 1. Create PeerConnection
const pc = new RTCPeerConnection({
  iceServers: [{urls: 'stun:stun.cloudflare.com:3478'}]
});

// 2. Add tracks
const stream = await navigator.mediaDevices.getUserMedia({video: true, audio: true});
stream.getTracks().forEach(track => pc.addTrack(track, stream));

// 3. Create offer
const offer = await pc.createOffer();
await pc.setLocalDescription(offer);

// 4. Send to backend → Cloudflare API
const response = await fetch('/api/new-session', {
  method: 'POST',
  body: JSON.stringify({sdp: offer.sdp})
});

// 5. Set remote answer
const {sessionDescription} = await response.json();
await pc.setRemoteDescription(sessionDescription);
```

## Publishing

```typescript
const offer = await pc.createOffer();
await pc.setLocalDescription(offer);

const res = await fetch(`/api/sessions/${sessionId}/tracks`, {
  method: 'POST',
  body: JSON.stringify({
    sdp: offer.sdp,
    tracks: [{location: 'local', trackName: 'my-video'}]
  })
});

const {sessionDescription, tracks} = await res.json();
await pc.setRemoteDescription(sessionDescription);
const publishedTrackId = tracks[0].trackName; // Share with others
```

## Subscribing

```typescript
const res = await fetch(`/api/sessions/${sessionId}/tracks`, {
  method: 'POST',
  body: JSON.stringify({
    tracks: [{location: 'remote', trackName: remoteTrackId, sessionId: remoteSessionId}]
  })
});

const {sessionDescription} = await res.json();
await pc.setRemoteDescription(sessionDescription);

const answer = await pc.createAnswer();
await pc.setLocalDescription(answer);

await fetch(`/api/sessions/${sessionId}/renegotiate`, {
  method: 'PUT',
  body: JSON.stringify({sdp: answer.sdp})
});

pc.ontrack = (event) => {
  const [remoteStream] = event.streams;
  videoElement.srcObject = remoteStream;
};
```

# Configuration & Deployment

## Dashboard Setup

1. Navigate to https://dash.cloudflare.com/?to=/:account/calls
2. Click "Create Application" (or use existing app)
3. Copy `CALLS_APP_ID` from dashboard
4. Generate and copy `CALLS_APP_SECRET` (treat as sensitive credential)
5. Use credentials in Wrangler config or environment variables below

## Dependencies

**Backend (Workers):** Built-in fetch API, no additional packages required

**Client (PartyTracks):**
```bash
npm install partytracks @cloudflare/calls
```

**Client (React + PartyTracks):**
```bash
npm install partytracks @cloudflare/calls observable-hooks
# Observable hooks: useObservableAsValue, useValueAsObservable
```

**Client (Raw API):** Native browser WebRTC API only

## Wrangler Setup

```jsonc
{
  "name": "my-calls-app",
  "main": "src/index.ts",
  "compatibility_date": "2025-01-01", // Use current date for new projects
  "vars": {
    "CALLS_APP_ID": "your-app-id",
    "MAX_WEBCAM_BITRATE": "1200000",
    "MAX_WEBCAM_FRAMERATE": "24",
    "MAX_WEBCAM_QUALITY_LEVEL": "1080"
  },
  // Set secret: wrangler secret put CALLS_APP_SECRET
  "durable_objects": {
    "bindings": [
      {
        "name": "ROOM",
        "class_name": "Room"
      }
    ]
  }
}
```

## Deploy

```bash
wrangler login
wrangler secret put CALLS_APP_SECRET
wrangler deploy
```

## Environment Variables

**Required:**
- `CALLS_APP_ID`: From dashboard
- `CALLS_APP_SECRET`: From dashboard (secret)

**Optional:**
- `MAX_WEBCAM_BITRATE` (default: 1200000)
- `MAX_WEBCAM_FRAMERATE` (default: 24)
- `MAX_WEBCAM_QUALITY_LEVEL` (default: 1080)
- `TURN_SERVICE_ID`: TURN service
- `TURN_SERVICE_TOKEN`: TURN auth (secret)

## TURN Configuration

```javascript
const pc = new RTCPeerConnection({
  iceServers: [
    { urls: 'stun:stun.cloudflare.com:3478' },
    {
      urls: [
        'turn:turn.cloudflare.com:3478?transport=udp',
        'turn:turn.cloudflare.com:3478?transport=tcp',
        'turns:turn.cloudflare.com:5349?transport=tcp'
      ],
      username: turnUsername,
      credential: turnCredential
    }
  ],
  bundlePolicy: 'max-bundle', // Recommended: reduces overhead
  iceTransportPolicy: 'all'    // Use 'relay' to force TURN (testing only)
});
```

**Ports:** 3478 (UDP/TCP), 53 (UDP), 80 (TCP), 443 (TLS), 5349 (TLS)

**When to use TURN:** Required for restrictive corporate firewalls/networks that block UDP. ~5-10% of connections fallback to TURN. STUN works for most users.

**ICE candidate filtering:** Cloudflare handles candidate filtering automatically. No need to manually filter candidates.

## Durable Object Boilerplate

Minimal presence system:

```typescript
export class Room {
  private sessions = new Map<string, {userId: string, tracks: string[]}>();

  async fetch(req: Request) {
    const {pathname} = new URL(req.url);
    const body = await req.json();
    
    if (pathname === '/join') {
      this.sessions.set(body.sessionId, {userId: body.userId, tracks: []});
      return Response.json({participants: this.sessions.size});
    }
    
    if (pathname === '/publish') {
      this.sessions.get(body.sessionId)?.tracks.push(...body.tracks);
      // Broadcast to others via WebSocket (not shown)
      return new Response('OK');
    }
    
    return new Response('Not found', {status: 404});
  }
}
```

## Environment Validation

Check credentials before first API call:

```typescript
if (!env.CALLS_APP_ID || !env.CALLS_APP_SECRET) {
  throw new Error('CALLS_APP_ID and CALLS_APP_SECRET required');
}
```

# Gotchas & Troubleshooting

## Common Errors

### "Slow initial connect (~1.8s)"

**Cause:** First STUN delayed during consensus forming (normal behavior)
**Solution:** Subsequent connections are faster. CF detects DTLS ClientHello early to compensate.

### "No media flow"

**Cause:** SDP exchange incomplete, connection not established, tracks not added before offer, browser permissions missing
**Solution:** 
1. Verify SDP exchange complete
2. Check `pc.connectionState === 'connected'`
3. Ensure tracks added before creating offer
4. Confirm browser permissions granted
5. Use `chrome://webrtc-internals` for debugging

### "Track not receiving"

**Cause:** Track not published, track ID not shared, session IDs mismatch, `pc.ontrack` not set, renegotiation needed
**Solution:** 
1. Verify track published successfully
2. Confirm track ID shared between peers
3. Check session IDs match
4. Set `pc.ontrack` handler before answer
5. Trigger renegotiation if needed

### "ICE connection failed"

**Cause:** Network changed, firewall blocked UDP, TURN needed, transient network issue
**Solution:**
```typescript
pc.oniceconnectionstatechange = async () => {
  if (pc.iceConnectionState === 'failed') {
    console.warn('ICE failed, attempting restart');
    await pc.restartIce(); // Triggers new ICE gathering
    
    // Create new offer with ICE restart flag
    const offer = await pc.createOffer({iceRestart: true});
    await pc.setLocalDescription(offer);
    
    // Send to backend → Cloudflare API
    await fetch(`/api/sessions/${sessionId}/renegotiate`, {
      method: 'PUT',
      body: JSON.stringify({sdp: offer.sdp})
    });
  }
};
```

### "Track stuck/frozen"

**Cause:** Sender paused track, network congestion, codec mismatch, mobile browser backgrounded
**Solution:**
1. Check `track.enabled` and `track.readyState === 'live'`
2. Verify sender active: `pc.getSenders().find(s => s.track === track)`
3. Check stats for packet loss/jitter (see patterns.md)
4. On mobile: Re-acquire tracks when app foregrounded
5. Test with different codecs if persistent

### "Network change disconnects call"

**Cause:** Mobile switching WiFi↔cellular, laptop changing networks
**Solution:**
```typescript
// Listen for network changes
if ('connection' in navigator) {
  (navigator as any).connection.addEventListener('change', async () => {
    console.log('Network changed');
    await pc.restartIce(); // Use ICE restart pattern above
  });
}

// Or use PartyTracks (handles automatically)
```

## Retry with Exponential Backoff

```typescript
async function fetchWithRetry(url: string, options: RequestInit, maxRetries = 3) {
  for (let i = 0; i < maxRetries; i++) {
    try {
      const res = await fetch(url, options);
      if (res.ok) return res;
      if (res.status >= 500) throw new Error('Server error');
      return res; // Client error, don't retry
    } catch (err) {
      if (i === maxRetries - 1) throw err;
      const delay = Math.min(1000 * 2 ** i, 10000); // Cap at 10s
      await new Promise(resolve => setTimeout(resolve, delay));
    }
  }
}
```

## Debugging with chrome://webrtc-internals

1. Open `chrome://webrtc-internals` in Chrome/Edge
2. Find your PeerConnection in the list
3. Check **Stats graphs** for packet loss, jitter, bandwidth
4. Check **ICE candidate pairs**: Look for `succeeded` state, relay vs host candidates
5. Check **getStats**: Raw metrics for inbound/outbound RTP
6. Look for errors in **Event log**: `iceConnectionState`, `connectionState` changes
7. Export data with "Download the PeerConnection updates and stats data" button
8. Common issues visible here: ICE failures, high packet loss, bitrate drops

## Limits

| Resource/Limit | Value | Notes |
|----------------|-------|-------|
| Egress (Free) | 1TB/month | Per account |
| Egress (Paid) | $0.05/GB | After free tier |
| Inbound traffic | Free | All plans |
| TURN service | Free | Included with SFU |
| Participants | No hard limit | Client bandwidth/CPU bound (typically 10-50 tracks) |
| Tracks per session | No hard limit | Client resources limited |
| Session duration | No hard limit | Production calls run for hours |
| WebRTC ports | UDP 1024-65535 | Outbound only, required for media |
| API rate limit | 600 req/min | Per app, burst allowed |

## Security Checklist

- ✅ **Never expose** `CALLS_APP_SECRET` to client
- ✅ **Validate user identity** in backend before creating sessions
- ✅ **Implement auth tokens** for session access (JWT in custom header)
- ✅ **Rate limit** session creation endpoints
- ✅ **Expire sessions** server-side after inactivity
- ✅ **Validate track IDs** before subscribing (prevent unauthorized access)
- ✅ **Use HTTPS** for all signaling (API calls)
- ✅ **Enable DTLS-SRTP** (automatic with Cloudflare, encrypts media)
- ⚠️ **Consider E2EE** for sensitive content (implement client-side with Insertable Streams API)

# Patterns & Use Cases

## Architecture

```
Client (WebRTC) <---> CF Edge <---> Backend (HTTP)
                           |
                    CF Backbone (310+ DCs)
                           |
                    Other Edges <---> Other Clients
```

Anycast: Last-mile <50ms (95%), no region select, NACK shield, distributed consensus

Cascading trees auto-scale to millions:
```
Publisher -> Edge A -> Edge B -> Sub1
                    \-> Edge C -> Sub2,3
```

## Use Cases

**1:1:** A creates session+publishes, B creates+subscribes to A+publishes, A subscribes to B
**N:N:** All create session+publish, backend broadcasts track IDs, all subscribe to others
**1:N:** Publisher creates+publishes, viewers each create+subscribe (no fan-out limit)
**Breakout:** Same PeerConnection! Backend closes/adds tracks, no recreation

## PartyTracks (Recommended)

Observable-based client with automatic device/network handling:

```typescript
import {PartyTracks} from 'partytracks';

// Create client
const pt = new PartyTracks({
  apiUrl: '/api/calls',
  sessionId: 'my-session',
  onTrack: (track, peer) => {
    const video = document.getElementById(`video-${peer.id}`) as HTMLVideoElement;
    video.srcObject = new MediaStream([track]);
  }
});

// Publish camera (push API)
const camera = await pt.getCamera(); // Auto-requests permissions, handles device changes
await pt.publishTrack(camera, {trackName: 'my-camera'});

// Subscribe to remote track (pull API)
await pt.subscribeToTrack({trackName: 'remote-camera', sessionId: 'other-session'});

// React hook example
import {useObservableAsValue} from 'observable-hooks';

function VideoCall() {
  const localTracks = useObservableAsValue(pt.localTracks$);
  const remoteTracks = useObservableAsValue(pt.remoteTracks$);
  
  return <div>{/* Render tracks */}</div>;
}

// Screenshare
const screen = await pt.getScreenshare();
await pt.publishTrack(screen, {trackName: 'my-screen'});

// Handle device changes (automatic)
// PartyTracks detects device changes (e.g., Bluetooth headset) and renegotiates
```

## Backend

Express:
```js
app.post('/api/new-session', async (req, res) => {
  const r = await fetch(`${CALLS_API}/apps/${process.env.CALLS_APP_ID}/sessions/new`,
    {method: 'POST', headers: {'Authorization': `Bearer ${process.env.CALLS_APP_SECRET}`}});
  res.json(await r.json());
});
```

Workers: Same pattern, use `env.CALLS_APP_ID` and `env.CALLS_APP_SECRET`

DO Presence: See configuration.md for boilerplate

## Audio Level Detection

```typescript
// Attach analyzer to audio track
function attachAudioLevelDetector(track: MediaStreamTrack) {
  const ctx = new AudioContext();
  const analyzer = ctx.createAnalyser();
  const src = ctx.createMediaStreamSource(new MediaStream([track]));
  src.connect(analyzer);
  
  const data = new Uint8Array(analyzer.frequencyBinCount);
  const checkLevel = () => {
    analyzer.getByteFrequencyData(data);
    const level = data.reduce((a, b) => a + b) / data.length;
    if (level > 30) console.log('Speaking:', level); // Trigger UI update
    requestAnimationFrame(checkLevel);
  };
  checkLevel();
}
```

## Connection Quality Monitoring

```typescript
pc.getStats().then(stats => {
  stats.forEach(report => {
    if (report.type === 'inbound-rtp' && report.kind === 'video') {
      const {packetsLost, packetsReceived, jitter} = report;
      const lossRate = packetsLost / (packetsLost + packetsReceived);
      if (lossRate > 0.05) console.warn('High packet loss:', lossRate);
      if (jitter > 100) console.warn('High jitter:', jitter);
    }
  });
});
```

## Stage Management (Limit Visible Participants)

```typescript
// Subscribe to top 6 active speakers only
let activeSubscriptions = new Set<string>();

function updateStage(topSpeakers: string[]) {
  const toAdd = topSpeakers.filter(id => !activeSubscriptions.has(id)).slice(0, 6);
  const toRemove = [...activeSubscriptions].filter(id => !topSpeakers.includes(id));
  
  toRemove.forEach(id => {
    pc.getSenders().find(s => s.track?.id === id)?.track?.stop();
    activeSubscriptions.delete(id);
  });
  
  toAdd.forEach(async id => {
    await fetch(`/api/subscribe`, {method: 'POST', body: JSON.stringify({trackId: id})});
    activeSubscriptions.add(id);
  });
}
```

## Advanced

Bandwidth mgmt:
```ts
const s = pc.getSenders().find(s => s.track?.kind === 'video');
const p = s.getParameters();
if (!p.encodings) p.encodings = [{}];
p.encodings[0].maxBitrate = 1200000; p.encodings[0].maxFramerate = 24;
await s.setParameters(p);
```

Simulcast (CF auto-forwards best layer):
```ts
pc.addTransceiver('video', {direction: 'sendonly', sendEncodings: [
  {rid: 'high', maxBitrate: 1200000},
  {rid: 'med', maxBitrate: 600000, scaleResolutionDownBy: 2},
  {rid: 'low', maxBitrate: 200000, scaleResolutionDownBy: 4}
]});
```

DataChannel:
```ts
const dc = pc.createDataChannel('chat', {ordered: true, maxRetransmits: 3});
dc.onopen = () => dc.send(JSON.stringify({type: 'chat', text: 'Hi'}));
dc.onmessage = (e) => console.log('RX:', JSON.parse(e.data));
```

**WHIP/WHEP:** For streaming interop (OBS → SFU, SFU → video players), use WHIP (ingest) and WHEP (egress) protocols. See Cloudflare Stream integration docs.

Integrations: R2 for recording `env.R2_BUCKET.put(...)`, Queues for analytics

Perf: 100-250ms connect, ~50ms latency (95%), 200-400ms glass-to-glass, no participant limit (client: 10-50 tracks)

# Cloudflare Realtime SFU Reference

Expert guidance for building real-time audio/video/data applications using Cloudflare Realtime SFU (Selective Forwarding Unit).

## Reading Order

| Task | Files | ~Tokens |
|------|-------|---------|
| New project | README → configuration | ~1200 |
| Implement publish/subscribe | README → api | ~1600 |
| Add PartyTracks | patterns (PartyTracks section) | ~800 |
| Build presence system | patterns (DO section) | ~800 |
| Debug connection issues | gotchas | ~700 |
| Scale to millions | patterns (Cascading section) | ~600 |
| Add simulcast | patterns (Advanced section) | ~500 |
| Configure TURN | configuration (TURN section) | ~400 |

## In This Reference

- **[configuration.md](configuration.md)** - Setup, deployment, environment variables, Wrangler config
- **[api.md](api.md)** - Sessions, tracks, endpoints, request/response patterns
- **[patterns.md](patterns.md)** - Architecture patterns, use cases, integration examples
- **[gotchas.md](gotchas.md)** - Common issues, debugging, performance, security

## Quick Start

Cloudflare Realtime SFU: WebRTC infrastructure on global network (310+ cities). Anycast routing, no regional constraints, pub/sub model.

**Core concepts:**
- **Sessions:** WebRTC PeerConnection to Cloudflare edge
- **Tracks:** Audio/video/data channels you publish or subscribe to
- **No rooms:** Build presence layer yourself via track sharing (see patterns.md)

**Mental model:** Your client establishes one WebRTC session, publishes tracks (audio/video), shares track IDs via your backend, others subscribe to your tracks using track IDs + your session ID.

## Choose Your Approach

| Approach | When to Use | Complexity |
|----------|-------------|------------|
| **PartyTracks** | Production apps with device switching, React | Low - Observable-based, handles reconnections |
| **Raw API** | Custom requirements, non-browser, learning | Medium - Full control, manual WebRTC lifecycle |
| **RealtimeKit** | End-to-end SDK with UI components | Lowest - Managed state, React hooks |

**Recommendation:** Start with PartyTracks for most production applications. See patterns.md for PartyTracks examples.

## SFU vs RealtimeKit

- **Realtime SFU:** WebRTC infrastructure (this reference). Build your own signaling, presence, UI.
- **RealtimeKit:** SDK layer on top of SFU. Includes React hooks, state management, UI components. Part of Cloudflare AI platform.

Use SFU directly when you need custom signaling or non-React framework. Use RealtimeKit for faster development with React.

## Setup

Dashboard: https://dash.cloudflare.com/?to=/:account/calls

Get `CALLS_APP_ID` and `CALLS_APP_SECRET` from dashboard, then see configuration.md for deployment.

## See Also

- [Orange Meets Demo](https://demo.orange.cloudflare.dev/)
- [Orange Source](https://github.com/cloudflare/orange)
- [Calls Examples](https://github.com/cloudflare/calls-examples)
- [API Reference](https://developers.cloudflare.com/api/resources/calls/)
- [RealtimeKit Docs](https://developers.cloudflare.com/realtime/realtimekit/)

# RealtimeKit API Reference

Complete API reference for Meeting object, REST endpoints, and SDK methods.

## Meeting Object API

### `meeting.self` - Local Participant

```typescript
// Properties: id, userId, name, audioEnabled, videoEnabled, screenShareEnabled, audioTrack, videoTrack, screenShareTracks, roomJoined, roomState
// Methods
await meeting.self.enableAudio() / disableAudio() / enableVideo() / disableVideo() / enableScreenShare() / disableScreenShare()
await meeting.self.setName("Name")  // Before join only
await meeting.self.setDevice(device)
const devices = await meeting.self.getAllDevices() / getAudioDevices() / getVideoDevices() / getSpeakerDevices()
// Events: 'roomJoined', 'audioUpdate', 'videoUpdate', 'screenShareUpdate', 'deviceUpdate', 'deviceListUpdate'
meeting.self.on('roomJoined', () => {})
meeting.self.on('audioUpdate', ({ audioEnabled, audioTrack }) => {})
```

### `meeting.participants` - Remote Participants

**Collections**:
```typescript
meeting.participants.joined / active / waitlisted / pinned  // Maps
const participants = meeting.participants.joined.toArray()
const count = meeting.participants.joined.size()
const p = meeting.participants.joined.get('peer-id')
```

**Participant Properties**:
```typescript
participant.id / userId / name
participant.audioEnabled / videoEnabled / screenShareEnabled
participant.audioTrack / videoTrack / screenShareTracks
```

**Events**:
```typescript
meeting.participants.joined.on('participantJoined', (participant) => {})
meeting.participants.joined.on('participantLeft', (participant) => {})
```

### `meeting.meta` - Metadata
```typescript
meeting.meta.meetingId / meetingTitle / meetingStartedTimestamp
```

### `meeting.chat` - Chat
```typescript
meeting.chat.messages  // Array
await meeting.chat.sendTextMessage("Hello") / sendImageMessage(file)
meeting.chat.on('chatUpdate', ({ message, messages }) => {})
```

### `meeting.polls` - Polling
```typescript
meeting.polls.items  // Array
await meeting.polls.create(question, options, anonymous, hideVotes)
await meeting.polls.vote(pollId, optionIndex)
```

### `meeting.plugins` - Collaborative Apps
```typescript
meeting.plugins.all  // Array
await meeting.plugins.activate(pluginId) / deactivate()
```

### `meeting.ai` - AI Features
```typescript
meeting.ai.transcripts  // Live transcriptions (when enabled in Preset)
```

### Core Methods
```typescript
await meeting.join()   // Emits 'roomJoined' on meeting.self
await meeting.leave()
```

## TypeScript Types

```typescript
import type { RealtimeKitClient, States, UIConfig, Participant } from '@cloudflare/realtimekit';

// Main interface
interface RealtimeKitClient {
  self: SelfState;          // Local participant (id, userId, name, audioEnabled, videoEnabled, roomJoined, roomState)
  participants: { joined, active, waitlisted, pinned };  // Reactive Maps
  chat: ChatNamespace;      // messages[], sendTextMessage(), sendImageMessage()
  polls: PollsNamespace;    // items[], create(), vote()
  plugins: PluginsNamespace;  // all[], activate(), deactivate()
  ai: AINamespace;          // transcripts[]
  meta: MetaState;          // meetingId, meetingTitle, meetingStartedTimestamp
  join(): Promise<void>;
  leave(): Promise<void>;
}

// Participant (self & remote share same shape)
interface Participant {
  id: string;                      // Peer ID (changes on rejoin)
  userId: string;                  // Persistent participant ID
  name: string;
  audioEnabled: boolean;
  videoEnabled: boolean;
  screenShareEnabled: boolean;
  audioTrack: MediaStreamTrack | null;
  videoTrack: MediaStreamTrack | null;
  screenShareTracks: MediaStreamTrack[];
}
```

## Store Architecture

RealtimeKit uses reactive store (event-driven updates, live Maps):

```typescript
// Subscribe to state changes
meeting.self.on('audioUpdate', ({ audioEnabled, audioTrack }) => {});
meeting.participants.joined.on('participantJoined', (p) => {});

// Access current state synchronously
const isAudioOn = meeting.self.audioEnabled;
const count = meeting.participants.joined.size();
```

**Key principles:** State updates emit events after changes. Use `.toArray()` sparingly. Collections are live Maps.

## REST API

Base: `https://api.cloudflare.com/client/v4/accounts/{account_id}/realtime/kit/{app_id}`

### Meetings
```bash
GET    /meetings                                    # List all
GET    /meetings/{meeting_id}                       # Get details
POST   /meetings                                    # Create: {"title": "..."}
PATCH  /meetings/{meeting_id}                       # Update: {"title": "...", "record_on_start": true}
```

### Participants
```bash
GET    /meetings/{meeting_id}/participants                          # List all
GET    /meetings/{meeting_id}/participants/{participant_id}         # Get details
POST   /meetings/{meeting_id}/participants                          # Add: {"name": "...", "preset_name": "...", "custom_participant_id": "..."}
PATCH  /meetings/{meeting_id}/participants/{participant_id}         # Update: {"name": "...", "preset_name": "..."}
DELETE /meetings/{meeting_id}/participants/{participant_id}         # Delete
POST   /meetings/{meeting_id}/participants/{participant_id}/token   # Refresh token
```

### Active Session
```bash
GET  /meetings/{meeting_id}/active-session               # Get active session
POST /meetings/{meeting_id}/active-session/kick          # Kick users: {"user_ids": ["id1", "id2"]}
POST /meetings/{meeting_id}/active-session/kick-all      # Kick all
POST /meetings/{meeting_id}/active-session/poll          # Create poll: {"question": "...", "options": [...], "anonymous": false}
```

### Recording
```bash
GET  /recordings?meeting_id={meeting_id}                 # List recordings
GET  /recordings/active-recording/{meeting_id}           # Get active recording
POST /recordings                                         # Start: {"meeting_id": "...", "type": "composite"} (or "track")
PUT  /recordings/{recording_id}                          # Control: {"action": "pause"} (or "resume", "stop")
POST /recordings/track                                   # Track recording: {"meeting_id": "...", "layers": [...]}
```

### Livestreaming
```bash
GET  /livestreams?exclude_meetings=false                                # List all
GET  /livestreams/{livestream_id}                                       # Get details
POST /meetings/{meeting_id}/livestreams                                 # Start for meeting
POST /meetings/{meeting_id}/active-livestream/stop                      # Stop
POST /livestreams                                                       # Create independent: returns {ingest_server, stream_key, playback_url}
```

### Sessions & Analytics
```bash
GET  /sessions                                                          # List all
GET  /sessions/{session_id}                                             # Get details
GET  /sessions/{session_id}/participants                                # List participants
GET  /sessions/{session_id}/participants/{participant_id}               # Call stats
GET  /sessions/{session_id}/chat                                        # Download chat CSV
GET  /sessions/{session_id}/transcript                                  # Download transcript CSV
GET  /sessions/{session_id}/summary                                     # Get summary
POST /sessions/{session_id}/summary                                     # Generate summary
GET  /analytics/daywise?start_date=YYYY-MM-DD&end_date=YYYY-MM-DD      # Day-wise analytics
GET  /analytics/livestreams/overall                                     # Livestream analytics
```

### Webhooks
```bash
GET    /webhooks                    # List all
POST   /webhooks                    # Create: {"url": "https://...", "events": ["session.started", "session.ended"]}
PATCH  /webhooks/{webhook_id}       # Update
DELETE /webhooks/{webhook_id}       # Delete
```

## Session Lifecycle

```
Initialization → Join Intent → [Waitlist?] → Meeting Screen (Stage) → Ended
                                   ↓ Approved
                               [Rejected → Ended]
```

UI Kit handles state transitions automatically.

## See Also

- [Configuration](./configuration.md) - Setup and installation
- [Patterns](./patterns.md) - Usage examples
- [README](./README.md) - Overview and quick start

# RealtimeKit Configuration

Configuration guide for RealtimeKit setup, client SDKs, and wrangler integration.

## Installation

### React
```bash
npm install @cloudflare/realtimekit @cloudflare/realtimekit-react-ui
```

### Angular
```bash
npm install @cloudflare/realtimekit @cloudflare/realtimekit-angular-ui
```

### Web Components/HTML
```bash
npm install @cloudflare/realtimekit @cloudflare/realtimekit-ui
```

## Client SDK Configuration

### React UI Kit
```tsx
import { RtkMeeting } from '@cloudflare/realtimekit-react-ui';
<RtkMeeting authToken="<token>" onLeave={() => {}} />
```

### Angular UI Kit
```typescript
@Component({ template: `<rtk-meeting [authToken]="authToken" (rtkLeave)="onLeave($event)"></rtk-meeting>` })
export class AppComponent { authToken = '<token>'; onLeave() {} }
```

### Web Components
```html
<script type="module" src="https://cdn.jsdelivr.net/npm/@cloudflare/realtimekit-ui/dist/realtimekit-ui/realtimekit-ui.esm.js"></script>
<rtk-meeting id="meeting"></rtk-meeting>
<script>
  document.getElementById('meeting').authToken = '<token>';
</script>
```

### Core SDK Configuration
```typescript
import RealtimeKitClient from '@cloudflare/realtimekit';

const meeting = new RealtimeKitClient({
  authToken: '<token>',
  video: true, audio: true, autoSwitchAudioDevice: true,
  mediaConfiguration: {
    video: { width: { ideal: 1280 }, height: { ideal: 720 }, frameRate: { ideal: 30 } },
    audio: { echoCancellation: true, noiseSuppression: true, autoGainControl: true },
    screenshare: { width: { max: 1920 }, height: { max: 1080 }, frameRate: { ideal: 15 } }
  }
});
await meeting.join();
```

## Backend Setup

### Create App & Credentials

**Dashboard**: https://dash.cloudflare.com/?to=/:account/realtime/kit

**API**:
```bash
curl -X POST 'https://api.cloudflare.com/client/v4/accounts/<account_id>/realtime/kit/apps' \
  -H 'Content-Type: application/json' \
  -H 'Authorization: Bearer <api_token>' \
  -d '{"name": "My RealtimeKit App"}'
```

**Required Permissions**: API token with **Realtime / Realtime Admin** permissions

### Create Presets

```bash
curl -X POST 'https://api.cloudflare.com/client/v4/accounts/<account_id>/realtime/kit/<app_id>/presets' \
  -H 'Authorization: Bearer <api_token>' \
  -d '{
    "name": "host",
    "permissions": {
      "canShareAudio": true,
      "canShareVideo": true,
      "canRecord": true,
      "canLivestream": true,
      "canStartStopRecording": true
    }
  }'
```

## Wrangler Configuration

### Basic Configuration
```jsonc
// wrangler.jsonc
{
  "name": "realtimekit-app",
  "main": "src/index.ts",
  "compatibility_date": "2025-01-01",  // Use current date
  "vars": {
    "CLOUDFLARE_ACCOUNT_ID": "abc123",
    "REALTIMEKIT_APP_ID": "xyz789"
  }
  // Secrets: wrangler secret put CLOUDFLARE_API_TOKEN
}
```

### With Database & Storage
```jsonc
{
  "d1_databases": [{ "binding": "DB", "database_name": "meetings", "database_id": "d1-id" }],
  "r2_buckets": [{ "binding": "RECORDINGS", "bucket_name": "recordings" }],
  "kv_namespaces": [{ "binding": "SESSIONS", "id": "kv-id" }]
}
```

### Multi-Environment
```bash
# Deploy to environments
wrangler deploy --env staging
wrangler deploy --env production
```

## TURN Service Configuration

RealtimeKit can use Cloudflare's TURN service for connectivity through restrictive networks:

```jsonc
// wrangler.jsonc
{
  "vars": {
    "TURN_SERVICE_ID": "your_turn_service_id"
  }
  // Set secret: wrangler secret put TURN_SERVICE_TOKEN
}
```

TURN automatically configured when enabled in account - no client-side changes needed.

## Theming & Design Tokens

```typescript
import type { UIConfig } from '@cloudflare/realtimekit';

const uiConfig: UIConfig = {
  designTokens: {
    colors: {
      brand: { 500: '#0066ff', 600: '#0052cc' },
      background: { 1000: '#1A1A1A', 900: '#2D2D2D' },
      text: { 1000: '#FFFFFF', 900: '#E0E0E0' }
    },
    borderRadius: 'extra-rounded',  // 'rounded' | 'extra-rounded' | 'sharp'
    theme: 'dark'  // 'light' | 'dark'
  },
  logo: { url: 'https://example.com/logo.png', altText: 'Company' }
};

// Apply to React
<RtkMeeting authToken={token} config={uiConfig} onLeave={() => {}} />

// Or use CSS variables
// :root { --rtk-color-brand-500: #0066ff; --rtk-border-radius: 12px; }
```

## Internationalization (i18n)

### Custom Language Strings
```typescript
import { useLanguage } from '@cloudflare/realtimekit-ui';

const customLanguage = {
  'join': 'Entrar',
  'leave': 'Salir',
  'mute': 'Silenciar',
  'unmute': 'Activar audio',
  'turn_on_camera': 'Encender cámara',
  'turn_off_camera': 'Apagar cámara',
  'share_screen': 'Compartir pantalla',
  'stop_sharing': 'Dejar de compartir'
};

const t = useLanguage(customLanguage);

// React usage
<RtkMeeting authToken={token} t={t} onLeave={() => {}} />
```

### Supported Locales
Default locales available: `en`, `es`, `fr`, `de`, `pt`, `ja`, `zh`

```typescript
import { setLocale } from '@cloudflare/realtimekit-ui';
setLocale('es');  // Switch to Spanish
```

## See Also

- [API](./api.md) - Meeting APIs, REST endpoints
- [Patterns](./patterns.md) - Backend integration examples
- [README](./README.md) - Overview and quick start

# RealtimeKit Gotchas & Troubleshooting

## Common Errors

### "Cannot connect to meeting"

**Cause:** Auth token invalid/expired, API credentials lack permissions, or network blocks WebRTC
**Solution:**
Verify token validity, check API token has **Realtime / Realtime Admin** permissions, enable TURN service for restrictive networks

### "No video/audio tracks"

**Cause:** Browser permissions not granted, video/audio not enabled, device in use, or device unavailable
**Solution:**
Request browser permissions explicitly, verify initialization config, use `meeting.self.getAllDevices()` to debug, close other apps using device

### "Participant count mismatched"

**Cause:** `meeting.participants` doesn't include `meeting.self`
**Solution:** Total count = `meeting.participants.joined.size() + 1`

### "Events not firing"

**Cause:** Listeners registered after actions, incorrect event name, or wrong namespace
**Solution:**
Register listeners before calling `meeting.join()`, check event names against docs, verify correct namespace

### "CORS errors in API calls"

**Cause:** Making REST API calls from client-side
**Solution:** All REST API calls **must** be server-side (Workers, backend). Never expose API tokens to clients.

### "Preset not applying"

**Cause:** Preset doesn't exist, name mismatch (case-sensitive), or participant created before preset
**Solution:**
Verify preset exists via Dashboard or API, check exact spelling and case, create preset before adding participants

### "Token reuse error"

**Cause:** Reusing participant tokens across sessions
**Solution:** Generate fresh token per session. Use refresh endpoint if token expires during session.

### "Video quality poor"

**Cause:** Insufficient bandwidth, resolution/bitrate too high, or CPU overload
**Solution:**
Lower `mediaConfiguration.video` resolution/frameRate, monitor network conditions, reduce participant count or grid size

### "Echo or audio feedback"

**Cause:** Multiple devices picking up same audio source
**Solution:**
- Lower `mediaConfiguration.video` resolution/frameRate
- Monitor network conditions
- Reduce participant count or grid size

### Issue: Echo or audio feedback
**Cause**: Multiple devices picking up same audio source

**Solutions**:
Enable `echoCancellation: true` in `mediaConfiguration.audio`, use headphones, mute when not speaking

### "Screen share not working"

**Cause:** Browser doesn't support screen sharing API, permission denied, or wrong `displaySurface` config
**Solution:**
Use Chrome/Edge/Firefox (Safari limited support), check browser permissions, try different `displaySurface` values ('window', 'monitor', 'browser')

### "How do I schedule meetings?"

**Cause:** RealtimeKit has no built-in scheduling system
**Solution:**
Store meeting IDs in your database with timestamps. Generate participant tokens only when user should join. Example:
```typescript
// Store in DB
{ meetingId: 'abc123', scheduledFor: '2026-02-15T10:00:00Z', userId: 'user456' }

// Generate token when user clicks "Join" near scheduled time
const response = await fetch('/api/join-meeting', {
  method: 'POST',
  body: JSON.stringify({ meetingId: 'abc123' })
});
const { authToken } = await response.json();
```

### "Recording not starting"

**Cause:** Preset lacks recording permissions, no active session, or API call from client
**Solution:**
Verify preset has `canRecord: true` and `canStartStopRecording: true`, ensure session is active (at least one participant), make recording API calls server-side only

## Limits

| Resource | Limit |
|----------|-------|
| Max participants per session | 100 |
| Max concurrent sessions per App | 1000 |
| Max recording duration | 6 hours |
| Max meeting duration | 24 hours |
| Max chat message length | 4000 characters |
| Max preset name length | 64 characters |
| Max meeting title length | 256 characters |
| Max participant name length | 256 characters |
| Token expiration | 24 hours (default) |
| WebRTC ports required | UDP 1024-65535 |

## Network Requirements

### Firewall Rules
Allow outbound UDP/TCP to:
- `*.cloudflare.com` ports 443, 80
- UDP ports 1024-65535 (WebRTC media)

### TURN Service
Enable for users behind restrictive firewalls/proxies:
```jsonc
// wrangler.jsonc
{
  "vars": {
    "TURN_SERVICE_ID": "your_turn_service_id"
  }
  // Set secret: wrangler secret put TURN_SERVICE_TOKEN
}
```

TURN automatically configured in SDK when enabled in account.

## Debugging Tips

```typescript
// Check devices
const devices = await meeting.self.getAllDevices();
meeting.self.on('deviceListUpdate', ({ added, removed, devices }) => console.log('Devices:', { added, removed, devices }));

// Monitor participants
meeting.participants.joined.on('participantJoined', (p) => console.log(`${p.name} joined:`, { id: p.id, userId: p.userId, audioEnabled: p.audioEnabled, videoEnabled: p.videoEnabled }));

// Check room state
meeting.self.on('roomJoined', () => console.log('Room:', { meetingId: meeting.meta.meetingId, meetingTitle: meeting.meta.meetingTitle, participantCount: meeting.participants.joined.size() + 1, audioEnabled: meeting.self.audioEnabled, videoEnabled: meeting.self.videoEnabled }));

// Log all events
['roomJoined', 'audioUpdate', 'videoUpdate', 'screenShareUpdate', 'deviceUpdate', 'deviceListUpdate'].forEach(event => meeting.self.on(event, (data) => console.log(`[self] ${event}:`, data)));
['participantJoined', 'participantLeft'].forEach(event => meeting.participants.joined.on(event, (data) => console.log(`[participants] ${event}:`, data)));
meeting.chat.on('chatUpdate', (data) => console.log('[chat] chatUpdate:', data));
```

## Security & Performance

### Security: Do NOT
- Expose `CLOUDFLARE_API_TOKEN` in client code, hardcode credentials in frontend
- Reuse participant tokens, store tokens in localStorage without encryption
- Allow client-side meeting creation

### Security: DO
- Generate tokens server-side only, use HTTPS, implement rate limiting
- Validate user auth before generating tokens, use `custom_participant_id` to map to your user system
- Set appropriate preset permissions per user role, rotate API tokens regularly

### Performance
- **CPU**: Lower video resolution/frameRate, disable video for audio-only, use `meeting.participants.active` for large meetings, implement virtual scrolling
- **Bandwidth**: Set max resolution in `mediaConfiguration`, disable screenshare audio if unneeded, use audio-only mode, implement adaptive bitrate
- **Memory**: Clean up event listeners on unmount, call `meeting.leave()` when done, don't store large participant arrays

## In This Reference
- [README.md](README.md) - Overview, core concepts, quick start
- [configuration.md](configuration.md) - SDK config, presets, wrangler setup
- [api.md](api.md) - Client SDK APIs, REST endpoints
- [patterns.md](patterns.md) - Common patterns, React hooks, backend integration

# RealtimeKit Patterns

## UI Kit (Minimal Code)

```tsx
// React
import { RtkMeeting } from '@cloudflare/realtimekit-react-ui';
<RtkMeeting authToken="<token>" onLeave={() => console.log('Left')} />

// Angular
@Component({ template: `<rtk-meeting [authToken]="authToken" (rtkLeave)="onLeave($event)"></rtk-meeting>` })
export class AppComponent { authToken = '<token>'; onLeave(event: unknown) {} }

// HTML/Web Components
<script type="module" src="https://cdn.jsdelivr.net/npm/@cloudflare/realtimekit-ui/dist/realtimekit-ui/realtimekit-ui.esm.js"></script>
<rtk-meeting id="meeting"></rtk-meeting>
<script>document.getElementById('meeting').authToken = '<token>';</script>
```

## UI Components

RealtimeKit provides 133+ pre-built Stencil.js Web Components with framework wrappers:

### Layout Components
- `<RtkMeeting>` - Full meeting UI (all-in-one)
- `<RtkHeader>`, `<RtkStage>`, `<RtkControlbar>` - Layout sections
- `<RtkSidebar>` - Chat/participants sidebar
- `<RtkGrid>` - Adaptive video grid

### Control Components  
- `<RtkMicToggle>`, `<RtkCameraToggle>` - Media controls
- `<RtkScreenShareToggle>` - Screen sharing
- `<RtkLeaveButton>` - Leave meeting
- `<RtkSettingsModal>` - Device settings

### Grid Variants
- `<RtkSpotlightGrid>` - Active speaker focus
- `<RtkAudioGrid>` - Audio-only mode
- `<RtkPaginatedGrid>` - Paginated layout

**See full catalog**: https://docs.realtime.cloudflare.com/ui-kit

## Core SDK Patterns

### Basic Setup
```typescript
import RealtimeKitClient from '@cloudflare/realtimekit';

const meeting = new RealtimeKitClient({ authToken, video: true, audio: true });
meeting.self.on('roomJoined', () => console.log('Joined:', meeting.meta.meetingTitle));
meeting.participants.joined.on('participantJoined', (p) => console.log(`${p.name} joined`));
await meeting.join();
```

### Video Grid & Device Selection
```typescript
// Video grid
function VideoGrid({ meeting }) {
  const [participants, setParticipants] = useState([]);
  useEffect(() => {
    const update = () => setParticipants(meeting.participants.joined.toArray());
    meeting.participants.joined.on('participantJoined', update);
    meeting.participants.joined.on('participantLeft', update);
    update();
    return () => { meeting.participants.joined.off('participantJoined', update); meeting.participants.joined.off('participantLeft', update); };
  }, [meeting]);
  return <div style={{ display: 'grid', gridTemplateColumns: 'repeat(auto-fill, minmax(300px, 1fr))' }}>
    {participants.map(p => <VideoTile key={p.id} participant={p} />)}
  </div>;
}

function VideoTile({ participant }) {
  const videoRef = useRef<HTMLVideoElement>(null);
  useEffect(() => {
    if (videoRef.current && participant.videoTrack) videoRef.current.srcObject = new MediaStream([participant.videoTrack]);
  }, [participant.videoTrack]);
  return <div><video ref={videoRef} autoPlay playsInline muted /><div>{participant.name}</div></div>;
}

// Device selection
const devices = await meeting.self.getAllDevices();
const switchCamera = (deviceId: string) => {
  const device = devices.find(d => d.deviceId === deviceId);
  if (device) await meeting.self.setDevice(device);
};
```

## React Hooks (Official)

```typescript
import { useRealtimeKitClient, useRealtimeKitSelector } from '@cloudflare/realtimekit-react-ui';

function MyComponent() {
  const [meeting, initMeeting] = useRealtimeKitClient();
  const audioEnabled = useRealtimeKitSelector(m => m.self.audioEnabled);
  const participantCount = useRealtimeKitSelector(m => m.participants.joined.size());
  
  useEffect(() => { initMeeting({ authToken: '<token>' }); }, []);
  
  return <div>
    <button onClick={() => meeting?.self.enableAudio()}>{audioEnabled ? 'Mute' : 'Unmute'}</button>
    <span>{participantCount} participants</span>
  </div>;
}
```

**Benefits:** Automatic re-renders, memoized selectors, type-safe

## Waitlist Handling

```typescript
// Monitor waitlist
meeting.participants.waitlisted.on('participantJoined', (participant) => {
  console.log(`${participant.name} is waiting`);
  // Show admin UI to approve/reject
});

// Approve from waitlist (backend only)
await fetch(
  `https://api.cloudflare.com/client/v4/accounts/${accountId}/realtime/kit/${appId}/meetings/${meetingId}/active-session/waitlist/approve`,
  {
    method: 'POST',
    headers: { 'Authorization': `Bearer ${apiToken}` },
    body: JSON.stringify({ user_ids: [participant.userId] })
  }
);

// Client receives automatic transition when approved
meeting.self.on('roomJoined', () => console.log('Approved and joined'));
```

## Audio-Only Mode

```typescript
const meeting = new RealtimeKitClient({
  authToken: '<token>',
  video: false,  // Disable video
  audio: true,
  mediaConfiguration: {
    audio: {
      echoCancellation: true,
      noiseSuppression: true,
      autoGainControl: true
    }
  }
});

// Use audio grid component
import { RtkAudioGrid } from '@cloudflare/realtimekit-react-ui';
<RtkAudioGrid meeting={meeting} />
```

## Addon System

```typescript
// List available addons
meeting.plugins.all.forEach(plugin => {
  console.log(plugin.id, plugin.name, plugin.active);
});

// Activate collaborative app
await meeting.plugins.activate('whiteboard-addon-id');

// Listen for activations
meeting.plugins.on('pluginActivated', ({ plugin }) => {
  console.log(`${plugin.name} activated`);
});

// Deactivate
await meeting.plugins.deactivate();
```

## Backend Integration

### Token Generation (Workers)
```typescript
export interface Env { CLOUDFLARE_API_TOKEN: string; CLOUDFLARE_ACCOUNT_ID: string; REALTIMEKIT_APP_ID: string; }

export default {
  async fetch(request: Request, env: Env): Promise<Response> {
    const url = new URL(request.url);
    
    if (url.pathname === '/api/join-meeting') {
      const { meetingId, userName, presetName } = await request.json();
      const response = await fetch(
        `https://api.cloudflare.com/client/v4/accounts/${env.CLOUDFLARE_ACCOUNT_ID}/realtime/kit/${env.REALTIMEKIT_APP_ID}/meetings/${meetingId}/participants`,
        {
          method: 'POST',
          headers: { 'Content-Type': 'application/json', 'Authorization': `Bearer ${env.CLOUDFLARE_API_TOKEN}` },
          body: JSON.stringify({ name: userName, preset_name: presetName })
        }
      );
      const data = await response.json();
      return Response.json({ authToken: data.result.authToken });
    }
    
    return new Response('Not found', { status: 404 });
  }
};
```

## Best Practices

### Security
1. **Never expose API tokens client-side** - Generate participant tokens server-side only
2. **Don't reuse participant tokens** - Generate fresh token per session, use refresh endpoint if expired
3. **Use custom participant IDs** - Map to your user system for cross-session tracking

### Performance
1. **Event-driven updates** - Listen to events, don't poll. Use `toArray()` only when needed
2. **Media quality constraints** - Set appropriate resolution/bitrate limits based on network conditions
3. **Device management** - Enable `autoSwitchAudioDevice` for better UX, handle device list updates

### Architecture
1. **Separate Apps for environments** - staging vs production to prevent data mixing
2. **Preset strategy** - Create presets at App level, reuse across meetings
3. **Token management** - Backend generates tokens, frontend receives via authenticated endpoint

## In This Reference
- [README.md](README.md) - Overview, core concepts, quick start
- [configuration.md](configuration.md) - SDK config, presets, wrangler setup
- [api.md](api.md) - Client SDK APIs, REST endpoints
- [gotchas.md](gotchas.md) - Common issues, troubleshooting, limits

# Cloudflare RealtimeKit

Expert guidance for building real-time video and audio applications using **Cloudflare RealtimeKit** - a comprehensive SDK suite for adding customizable live video and voice to web or mobile applications.

## Overview

RealtimeKit is Cloudflare's SDK suite built on Realtime SFU, abstracting WebRTC complexity with fast integration, pre-built UI components, global performance (300+ cities), and production features (recording, transcription, chat, polls).

**Use cases**: Team meetings, webinars, social video, audio calls, interactive plugins

## Core Concepts

- **App**: Workspace grouping meetings, participants, presets, recordings. Use separate Apps for staging/production
- **Meeting**: Re-usable virtual room. Each join creates new **Session**
- **Session**: Live meeting instance. Created on first join, ends after last leave
- **Participant**: User added via REST API. Returns `authToken` for client SDK. **Do not reuse tokens**
- **Preset**: Reusable permission/UI template (permissions, meeting type, theme). Applied at participant creation
- **Peer ID** (`id`): Unique per session, changes on rejoin
- **Participant ID** (`userId`): Persistent across sessions

## Quick Start

### 1. Create App & Meeting (Backend)

```bash
# Create app
curl -X POST 'https://api.cloudflare.com/client/v4/accounts/<account_id>/realtime/kit/apps' \
  -H 'Authorization: Bearer <api_token>' \
  -d '{"name": "My RealtimeKit App"}'

# Create meeting
curl -X POST 'https://api.cloudflare.com/client/v4/accounts/<account_id>/realtime/kit/<app_id>/meetings' \
  -H 'Authorization: Bearer <api_token>' \
  -d '{"title": "Team Standup"}'

# Add participant
curl -X POST 'https://api.cloudflare.com/client/v4/accounts/<account_id>/realtime/kit/<app_id>/meetings/<meeting_id>/participants' \
  -H 'Authorization: Bearer <api_token>' \
  -d '{"name": "Alice", "preset_name": "host"}'
# Returns: { authToken }
```

### 2. Client Integration

**React**:
```tsx
import { RtkMeeting } from '@cloudflare/realtimekit-react-ui';

function App() {
  return <RtkMeeting authToken="<participant_auth_token>" onLeave={() => {}} />;
}
```

**Core SDK**:
```typescript
import RealtimeKitClient from '@cloudflare/realtimekit';

const meeting = new RealtimeKitClient({ authToken: '<token>', video: true, audio: true });
await meeting.join();
```

## Reading Order

| Task | Files |
|------|-------|
| Quick integration | README only |
| Custom UI | README → patterns → api |
| Backend setup | README → configuration |
| Debug issues | gotchas |
| Advanced features | patterns → api |

## RealtimeKit vs Realtime SFU

| Choose | When |
|--------|------|
| **RealtimeKit** | Need pre-built UI, fast integration, React/Angular/HTML |
| **Realtime SFU** | Building from scratch, custom WebRTC, full control |

RealtimeKit is built on Realtime SFU but abstracts WebRTC complexity with UI components and SDKs.

## Which Package?

Need pre-built meeting UI?
- React → `@cloudflare/realtimekit-react-ui` (`<RtkMeeting>`)
- Angular → `@cloudflare/realtimekit-angular-ui`
- HTML/Vanilla → `@cloudflare/realtimekit-ui`

Need custom UI?
- Core SDK → `@cloudflare/realtimekit` (RealtimeKitClient) - full control

Need raw WebRTC control?
- See `realtime-sfu/` reference

## In This Reference

- [Configuration](./configuration.md) - Setup, installation, wrangler config
- [API](./api.md) - Meeting object, REST API, SDK methods
- [Patterns](./patterns.md) - Common workflows, code examples
- [Gotchas](./gotchas.md) - Common issues, troubleshooting

## See Also

- [Workers](https://developers.cloudflare.com/workers/) - Backend integration
- [D1](../d1/) - Meeting metadata storage
- [R2](../r2/) - Recording storage
- [KV](../kv/) - Session management

## Reference Links

- **Official Docs**: https://developers.cloudflare.com/realtime/realtimekit/
- **API Reference**: https://developers.cloudflare.com/api/resources/realtime_kit/
- **Examples**: https://github.com/cloudflare/realtimekit-web-examples
- **Dashboard**: https://dash.cloudflare.com/?to=/:account/realtime/kit

# Secrets Store APIs

Fetch the current API documentation before implementing calls or copying types.

| Task | Documentation |
|------|---------------|
| Read a bound account secret asynchronously in a Worker | [Workers integration: access the secret](https://developers.cloudflare.com/secrets-store/integrations/workers/#3-access-the-secret-on-the-env-object) |
| Generate binding and runtime types for the Worker configuration | [Workers TypeScript](https://developers.cloudflare.com/workers/languages/typescript/) |
| Manage stores and secrets, inspect metadata, or query quota | [Secrets Store REST API](https://developers.cloudflare.com/api/resources/secrets_store/) |
| Choose authorization and consuming-service scope | [Access control](https://developers.cloudflare.com/secrets-store/access-control/) |
| Manage secrets through the CLI instead of REST | [Wrangler Secrets Store commands](https://developers.cloudflare.com/workers/wrangler/commands/secrets-store/) |

## Runtime decisions

Account-secret bindings require asynchronous retrieval; they are not the direct string values exposed by [Worker secrets](https://developers.cloudflare.com/workers/configuration/secrets/). Account management APIs and metadata reads are separate from consuming a bound secret in a Worker.

Handle retrieval failures at the application's error boundary without exposing credentials. Reuse a retrieved value within the request when useful; avoid long-lived application caches that could keep revoked credentials in use. Validate structured secret values against the application's schema before using them.

Never return a credential to a client or include it in logs, error messages, or telemetry. Record only the non-sensitive context needed to diagnose a failure.

See [configuration.md](./configuration.md) for setup and [gotchas.md](./gotchas.md) for access failures.

# Secrets Store configuration

Fetch the guide for the operation you are performing before writing configuration or running management commands.

## Setup and management

| Task | Documentation |
|------|---------------|
| Create a store and secret, then bind it through Wrangler or the dashboard | [Workers integration](https://developers.cloudflare.com/secrets-store/integrations/workers/) |
| Create, edit, duplicate, or delete account secrets | [Manage secrets](https://developers.cloudflare.com/secrets-store/manage-secrets/how-to/) |
| Look up current store/secret command syntax and local versus remote flags | [Wrangler Secrets Store commands](https://developers.cloudflare.com/workers/wrangler/commands/secrets-store/) |
| Configure bindings for each deployment environment | [Wrangler environments](https://developers.cloudflare.com/workers/wrangler/environments/) |
| Choose user roles, CI token permissions, and secret scopes | [Secrets Store access control](https://developers.cloudflare.com/secrets-store/access-control/) |

Treat the store ID, secret ID, secret name, and Worker binding name as different identifiers. Use the identifier required by the documented operation; do not infer update or delete flags from the create command.

## Local development and deployment

Secrets Store management commands default to local state; production operations use the documented remote option. Local development needs separately provisioned local secrets. Follow the local-development notes in [Workers integration](https://developers.cloudflare.com/secrets-store/integrations/workers/) and the [command reference](https://developers.cloudflare.com/workers/wrangler/commands/secrets-store/).

Check the selected account, deployment environment, secret scope, and binding configuration before deploying. In CI, distinguish permission to read metadata from permission to attach a secret to a Worker; use the [CI/CD access-control guidance](https://developers.cloudflare.com/secrets-store/access-control/#api-token-permissions).

Use protected secret input rather than putting credential values in command arguments, source files, or CI logs. For interactive CLI use, follow the command reference's secret-value prompt guidance. Keep local credentials out of version control.

See [api.md](./api.md) for runtime access and [patterns.md](./patterns.md) before replacing a shared credential.

# Secrets Store troubleshooting

Start with the failing operation and fetch its documentation before changing credentials or bindings.

| Symptom or decision | What to check | Documentation |
|---------------------|---------------|---------------|
| Deployment cannot attach a secret | Selected account, caller's binding permission, and the secret's consuming-service scope; metadata read permission alone is insufficient | [Access control](https://developers.cloudflare.com/secrets-store/access-control/) |
| Secret is missing or the wrong value is used | Store, secret name, binding name, and selected deployment environment | [Workers integration](https://developers.cloudflare.com/secrets-store/integrations/workers/), [Wrangler environments](https://developers.cloudflare.com/workers/wrangler/environments/) |
| Secret works in production but fails locally | Local secret provisioning and the management command's local/remote target | [Workers integration](https://developers.cloudflare.com/secrets-store/integrations/workers/), [Wrangler commands](https://developers.cloudflare.com/workers/wrangler/commands/secrets-store/) |
| CLI update, retrieval, or deletion fails | The operation's required identifier and flags; a secret name is not interchangeable with its ID | [Wrangler commands](https://developers.cloudflare.com/workers/wrangler/commands/secrets-store/) |
| Binding is treated as a string or has incorrect types | Asynchronous account-secret access and generated configuration types | [Runtime access](https://developers.cloudflare.com/secrets-store/integrations/workers/#3-access-the-secret-on-the-env-object), [TypeScript](https://developers.cloudflare.com/workers/languages/typescript/) |
| Quota or value-size validation fails | Current account quota and the requested operation's schema | [Secrets Store REST API](https://developers.cloudflare.com/api/resources/secrets_store/) |
| Regional or integration support is unclear | Current product availability and supported consumers | [Product overview](https://developers.cloudflare.com/secrets-store/) |

Do not diagnose failures by printing secret values or returning raw errors to clients. Check metadata and sanitized operation context. Validate JSON or other structured values before consuming them, and handle retrieval or parsing failures at the application's error boundary.

Before deleting a secret to fix a binding conflict or quota problem, identify all consumers. Follow [secret management](https://developers.cloudflare.com/secrets-store/manage-secrets/how-to/) and the rotation decisions in [patterns.md](./patterns.md); removing a shared credential can affect multiple services.

See [configuration.md](./configuration.md) for setup and [api.md](./api.md) for API selection.

# Secrets Store patterns

Use these decisions to choose the relevant guides; fetch the linked documentation before implementing.

## Rotation and sharing

Fetch [edit, duplicate, and delete operations](https://developers.cloudflare.com/secrets-store/manage-secrets/how-to/) and [Workers binding setup](https://developers.cloudflare.com/secrets-store/integrations/workers/) before changing a credential. Editing a shared secret affects every service using it. Inventory consumers and coordinate the change with the credential's issuer.

An application rotation plan must account for old and new credential validity, consumer rollout, verification, rollback, and eventual revocation. The management guide describes secret operations, not an end-to-end zero-downtime rotation protocol. Do not retry arbitrary failed requests with an old key: retries must respect upstream authentication semantics and the operation's idempotency.

## Migrate from Worker secrets

Read [Worker secrets](https://developers.cloudflare.com/workers/configuration/secrets/) alongside [Secrets Store Workers integration](https://developers.cloudflare.com/secrets-store/integrations/workers/). Migration changes both the binding configuration and access from a direct value to asynchronous retrieval. Verify the new binding in staging, resolve naming conflicts during rollout, and remove the old secret only after consumers have switched successfully.

## Encryption and signing

For cryptographic operations, fetch [Workers Web Crypto](https://developers.cloudflare.com/workers/runtime-apis/web-crypto/) and the [request-signing example](https://developers.cloudflare.com/workers/examples/signing-requests/). Use the [Secrets Store integration](https://developers.cloudflare.com/secrets-store/integrations/workers/) to retrieve key material.

These pages cover the runtime primitives; they do not define an application's encrypted KV format, key lifecycle, or signing protocol. Choose those explicitly, including key encoding, nonce handling, verification, and rotation of data encrypted under old keys. Validate JSON secrets at runtime rather than relying on a TypeScript assertion.

## Audit and integrations

Fetch [Secrets Store audit logs](https://developers.cloudflare.com/secrets-store/audit-logs/) for the recorded actions and how to inspect them. Keep application telemetry free of credential values; do not assume account audit events replace application-level success and failure monitoring.

For AI Gateway credentials, use [Bring your own keys](https://developers.cloudflare.com/ai-gateway/configuration/bring-your-own-keys/). For an internal authentication service, consult [Worker service bindings](https://developers.cloudflare.com/workers/runtime-apis/bindings/service-bindings/) and define the application's authorization boundary before exposing signing or secret-backed operations.

See [configuration.md](./configuration.md) for permissions and environments, and [gotchas.md](./gotchas.md) for troubleshooting.

# Cloudflare Secrets Store

Use Secrets Store for account-level credentials shared across Workers or supported integrations. Use [Worker secrets](https://developers.cloudflare.com/workers/configuration/secrets/) when credentials belong to one Worker and do not need centralized sharing.

Fetch the relevant documentation before implementing. Current Cloudflare docs are the source of truth for binding APIs, management commands, permissions, availability, and quotas. Use the [Secrets Store documentation index](https://developers.cloudflare.com/secrets-store/llms.txt) to discover additional guidance.

## Choose the scope

- Share a secret only among services that should use the same credential and rotate together.
- Separate development, staging, and production credentials; select the intended account and environment before managing or binding a secret.
- Grant only the management permissions and consuming-service scopes needed. Permission to view metadata does not imply permission to bind or retrieve a value; fetch [access control](https://developers.cloudflare.com/secrets-store/access-control/) for the current rules.

## Read by task

| Task | Reference |
|------|-----------|
| Create secrets, configure bindings, or prepare local development | [configuration.md](./configuration.md) |
| Read a secret in a Worker or automate management | [api.md](./api.md) |
| Plan rotation, migration, encryption, or auditing | [patterns.md](./patterns.md) |
| Diagnose access, deployment, or quota failures | [gotchas.md](./gotchas.md) |

Fetch the [product overview](https://developers.cloudflare.com/secrets-store/) for current availability and supported integrations. For AI Gateway provider credentials, use [Bring your own keys](https://developers.cloudflare.com/ai-gateway/configuration/bring-your-own-keys/).

# Smart Placement API

## Placement Status API

Query Worker placement status via Cloudflare API:

```bash
curl -X GET "https://api.cloudflare.com/client/v4/accounts/{ACCOUNT_ID}/workers/services/{WORKER_NAME}" \
  -H "Authorization: Bearer <TOKEN>" \
  -H "Content-Type: application/json"
```

Response includes `placement_status` field:

```typescript
type PlacementStatus = 
  | undefined  // Not yet analyzed
  | 'SUCCESS'  // Successfully optimized
  | 'INSUFFICIENT_INVOCATIONS'  // Not enough traffic
  | 'UNSUPPORTED_APPLICATION';  // Made Worker slower (reverted)
```

## Status Meanings

**`undefined` (not present)**
- Worker not yet analyzed
- Always runs at default edge location closest to user

**`SUCCESS`**
- Analysis complete, Smart Placement active
- Worker runs in optimal location (may be edge or remote)

**`INSUFFICIENT_INVOCATIONS`**
- Not enough requests to make placement decision
- Requires consistent multi-region traffic
- Always runs at default edge location

**`UNSUPPORTED_APPLICATION`** (rare, <1% of Workers)
- Smart Placement made Worker slower
- Placement decision reverted
- Always runs at edge location
- Won't be re-analyzed until redeployed

## cf-placement Header (Beta)

Smart Placement adds response header indicating routing decision:

```typescript
// Remote placement (Smart Placement routed request)
"cf-placement: remote-LHR"  // Routed to London

// Local placement (default edge routing)  
"cf-placement: local-EWR"   // Stayed at Newark edge
```

Format: `{placement-type}-{IATA-code}`
- `remote-*` = Smart Placement routed to remote location
- `local-*` = Stayed at default edge location
- IATA code = nearest airport to data center

**Warning:** Beta feature, may be removed before GA.

## Detecting Smart Placement in Code

**Note:** `cf-placement` header is a beta feature and may change or be removed.

```typescript
export default {
  async fetch(request: Request, env: Env): Promise<Response> {
    const placementHeader = request.headers.get('cf-placement');
    
    if (placementHeader?.startsWith('remote-')) {
      const location = placementHeader.split('-')[1];
      console.log(`Smart Placement routed to ${location}`);
    } else if (placementHeader?.startsWith('local-')) {
      const location = placementHeader.split('-')[1];
      console.log(`Running at edge location ${location}`);
    }
    
    return new Response('OK');
  }
} satisfies ExportedHandler<Env>;
```

## Request Duration Metrics

Available in Cloudflare dashboard when Smart Placement enabled:

**Workers & Pages → [Your Worker] → Metrics → Request Duration**

Shows histogram comparing:
- Request duration WITH Smart Placement (99% of traffic)
- Request duration WITHOUT Smart Placement (1% baseline)

**Request Duration vs Execution Duration:**
- **Request duration:** Total time from request arrival to response delivery (includes network latency)
- **Execution duration:** Time Worker code actively executing (excludes network waits)

Use request duration to measure Smart Placement impact.

### Interpreting Metrics

| Metric Comparison | Interpretation | Action |
|-------------------|----------------|--------|
| WITH < WITHOUT | Smart Placement helping | Keep enabled |
| WITH ≈ WITHOUT | Neutral impact | Consider disabling to free resources |
| WITH > WITHOUT | Smart Placement hurting | Disable with `mode: "off"` |

**Why Smart Placement might hurt performance:**
- Worker primarily serves static assets or cached content
- Backend services are globally distributed (no single optimal location)
- Worker has minimal backend communication
- Using Pages with `assets.run_worker_first = true`

**Typical improvements when Smart Placement helps:**
- 20-50% reduction in request duration for database-heavy Workers
- 30-60% reduction for Workers making multiple backend API calls
- Larger improvements when backend is geographically concentrated

## Monitoring Commands

```bash
# Tail Worker logs
wrangler tail your-worker-name

# Tail with filters
wrangler tail your-worker-name --status error
wrangler tail your-worker-name --header cf-placement

# Check placement status via API
curl -H "Authorization: Bearer $TOKEN" \
  https://api.cloudflare.com/client/v4/accounts/$ACCOUNT_ID/workers/services/$WORKER_NAME \
  | jq .result.placement_status
```

## TypeScript Types

```typescript
// Placement status returned by API (field may be absent)
type PlacementStatus = 
  | 'SUCCESS'
  | 'INSUFFICIENT_INVOCATIONS'
  | 'UNSUPPORTED_APPLICATION'
  | undefined;

// Placement configuration in wrangler.jsonc
type PlacementMode = 'smart' | 'off';

interface PlacementConfig {
  mode: PlacementMode;
  // Legacy fields (deprecated/removed):
  // hint?: string;  // REMOVED - no longer supported
}

// Explicit placement (separate feature from Smart Placement)
interface ExplicitPlacementConfig {
  region?: string;
  host?: string;
  hostname?: string;
  // Cannot combine with mode field
}

// Worker metadata from API response
interface WorkerMetadata {
  placement?: PlacementConfig | ExplicitPlacementConfig;
  placement_status?: PlacementStatus;
}

// Service Binding for backend Worker
interface Env {
  BACKEND_SERVICE: Fetcher;  // Service Binding to backend Worker
  DATABASE: D1Database;
}

// Example Worker with Service Binding
export default {
  async fetch(request: Request, env: Env): Promise<Response> {
    // Forward to backend Worker with Smart Placement enabled
    const response = await env.BACKEND_SERVICE.fetch(request);
    return response;
  }
} satisfies ExportedHandler<Env>;
```

# Smart Placement Configuration

## wrangler.jsonc Setup

```jsonc
{
  "$schema": "./node_modules/wrangler/config-schema.json",
  "placement": {
    "mode": "smart"
  }
}
```

## Placement Mode Values

| Mode | Behavior |
|------|----------|
| `"smart"` | Enable Smart Placement - automatic optimization based on traffic analysis |
| `"off"` | Explicitly disable Smart Placement - always run at edge closest to user |
| Not specified | Default behavior - run at edge closest to user (same as `"off"`) |

**Note:** Smart Placement vs Explicit Placement are separate features. Smart Placement (`mode: "smart"`) uses automatic analysis. For manual placement control, see explicit placement options (`region`, `host`, `hostname` fields - not covered in this reference).

## Frontend + Backend Split Configuration

### Frontend Worker (No Smart Placement)

```jsonc
// frontend-worker/wrangler.jsonc
{
  "name": "frontend",
  "main": "frontend-worker.ts",
  // No "placement" - runs at edge
  "services": [
    {
      "binding": "BACKEND",
      "service": "backend-api"
    }
  ]
}
```

### Backend Worker (Smart Placement Enabled)

```jsonc
// backend-api/wrangler.jsonc
{
  "name": "backend-api",
  "main": "backend-worker.ts",
  "placement": {
    "mode": "smart"
  },
  "d1_databases": [
    {
      "binding": "DATABASE",
      "database_id": "xxx"
    }
  ]
}
```

## Requirements & Limitations

### Requirements
- **Wrangler version:** 2.20.0+
- **Analysis time:** Up to 15 minutes
- **Traffic requirements:** Consistent multi-location traffic
- **Workers plan:** All plans (Free, Paid, Enterprise)

### What Smart Placement Affects

**CRITICAL LIMITATION - Smart Placement ONLY Affects `fetch` Handlers:**

Smart Placement is fundamentally limited to Workers with default `fetch` handlers. This is a key architectural constraint.

- ✅ **Affects:** `fetch` event handlers ONLY (the default export's fetch method)
- ❌ **Does NOT affect:** 
  - RPC methods (Service Bindings with `WorkerEntrypoint` - see example below)
  - Named entrypoints (exports other than `default`)
  - Workers without `fetch` handlers
  - Queue consumers, scheduled handlers, or other event types

**Example - Smart Placement ONLY affects `fetch`:**
```typescript
// ✅ Smart Placement affects this:
export default {
  async fetch(request: Request, env: Env): Promise<Response> {
    // This runs close to backend when Smart Placement enabled
    const data = await env.DATABASE.prepare('SELECT * FROM users').all();
    return Response.json(data);
  }
}

// ❌ Smart Placement DOES NOT affect these:
export class MyRPC extends WorkerEntrypoint {
  async myMethod() { 
    // This ALWAYS runs at edge, Smart Placement has NO EFFECT
    const data = await this.env.DATABASE.prepare('SELECT * FROM users').all();
    return data;
  }
}

export async function scheduled(event: ScheduledEvent, env: Env) {
  // NOT affected by Smart Placement
}
```

**Consequence:** If your backend logic uses RPC methods (`WorkerEntrypoint`), Smart Placement cannot optimize those calls. You must use fetch-based patterns for Smart Placement to work.

**Solution:** Convert RPC methods to fetch endpoints, or use a wrapper Worker with `fetch` handler that calls your backend RPC (though this adds latency).

### Baseline Traffic
Smart Placement automatically routes 1% of requests WITHOUT optimization as baseline for performance comparison.

### Validation Rules

**Mutually exclusive fields:**
- `mode` cannot be used with explicit placement fields (`region`, `host`, `hostname`)
- Choose either Smart Placement OR explicit placement, not both

```jsonc
// ✅ Valid - Smart Placement
{ "placement": { "mode": "smart" } }

// ✅ Valid - Explicit Placement (different feature)
{ "placement": { "region": "us-east1" } }

// ❌ Invalid - Cannot combine
{ "placement": { "mode": "smart", "region": "us-east1" } }
```

## Dashboard Configuration

**Workers & Pages** → Select Worker → **Settings** → **General** → **Placement: Smart** → Wait 15min → Check **Metrics**

## TypeScript Types

```typescript
interface Env {
  BACKEND: Fetcher;
  DATABASE: D1Database;
}

export default {
  async fetch(request: Request, env: Env): Promise<Response> {
    const data = await env.DATABASE.prepare('SELECT * FROM table').all();
    return Response.json(data);
  }
} satisfies ExportedHandler<Env>;
```

## Cloudflare Pages/Assets Warning

**CRITICAL PERFORMANCE ISSUE:** Enabling Smart Placement with `assets.run_worker_first = true` in Pages projects **severely degrades asset serving performance**. This is one of the most common misconfigurations.

**Why this is bad:**
- Smart Placement routes ALL requests (including static assets) away from edge to remote locations
- Static assets (HTML, CSS, JS, images) should ALWAYS be served from edge closest to user
- Result: 2-5x slower asset loading times, poor user experience

**Problem:** Smart Placement routes asset requests away from edge, but static assets should always be served from edge closest to user.

**Solutions (in order of preference):**
1. **Recommended:** Split into separate Workers (frontend at edge + backend with Smart Placement)
2. Set `"mode": "off"` to explicitly disable Smart Placement for Pages/Assets Workers
3. Use `assets.run_worker_first = false` (serves assets first, bypasses Worker for static content)

```jsonc
// ❌ BAD - Degrades asset performance by 2-5x
{
  "name": "pages-app",
  "placement": { "mode": "smart" },
  "assets": { "run_worker_first": true }
}

// ✅ GOOD - Frontend at edge, backend optimized
// frontend-worker/wrangler.jsonc
{
  "name": "frontend",
  "assets": { "run_worker_first": true }
  // No placement - runs at edge
}

// backend-worker/wrangler.jsonc
{
  "name": "backend-api",
  "placement": { "mode": "smart" },
  "d1_databases": [{ "binding": "DB", "database_id": "xxx" }]
}
```

**Key takeaway:** Never enable Smart Placement on Workers that serve static assets with `run_worker_first = true`.

## Local Development

Smart Placement does NOT work in `wrangler dev` (local only). Test by deploying: `wrangler deploy --env staging`

# Smart Placement Gotchas

## Common Errors

### "INSUFFICIENT_INVOCATIONS"

**Cause:** Not enough traffic for Smart Placement to analyze
**Solution:**
- Ensure Worker receives consistent global traffic
- Wait longer (analysis takes up to 15 minutes)
- Send test traffic from multiple global locations
- Check Worker has fetch event handler

### "UNSUPPORTED_APPLICATION"

**Cause:** Smart Placement made Worker slower rather than faster
**Reasons:**
- Worker doesn't make backend calls (runs faster at edge)
- Backend calls are cached (network latency to user more important)
- Backend service has good global distribution
- Worker serves static assets or Pages content

**Solutions:**
- Disable Smart Placement: `{ "placement": { "mode": "off" } }`
- Review whether Worker actually benefits from Smart Placement
- Consider caching strategy to reduce backend calls
- For Pages/Assets Workers, use separate backend Worker with Smart Placement

### "No request duration metrics"

**Cause:** Smart Placement not enabled, insufficient time passed, insufficient traffic, or analysis incomplete
**Solution:**
- Ensure Smart Placement enabled in config
- Wait 15+ minutes after deployment
- Verify Worker has sufficient traffic
- Check `placement_status` is `SUCCESS`

### "cf-placement header missing"

**Cause:** Smart Placement not enabled, beta feature removed, or Worker not analyzed yet
**Solution:** Verify Smart Placement enabled, wait for analysis (15min), check if beta feature still available

## Pages/Assets + Smart Placement Performance Degradation

**Problem:** Static assets load 2-5x slower when Smart Placement enabled with `run_worker_first = true`.

**Cause:** Smart Placement routes ALL requests (including static assets like HTML, CSS, JS, images) to remote locations. Static content should ALWAYS be served from edge closest to user.

**Solution:** Split into separate Workers OR disable Smart Placement:
```jsonc
// ❌ BAD - Assets routed away from user
{
  "name": "pages-app",
  "placement": { "mode": "smart" },
  "assets": { "run_worker_first": true }
}

// ✅ GOOD - Assets at edge, API optimized
// frontend/wrangler.jsonc
{
  "name": "frontend",
  "assets": { "run_worker_first": true }
  // No placement field - stays at edge
}

// backend/wrangler.jsonc
{
  "name": "backend-api",
  "placement": { "mode": "smart" }
}
```

This is one of the most common and impactful Smart Placement misconfigurations.

## Monolithic Full-Stack Worker

**Problem:** Frontend and backend logic in single Worker with Smart Placement enabled.

**Cause:** Smart Placement optimizes for backend latency but increases user-facing response time.

**Solution:** Split into two Workers:
```jsonc
// frontend/wrangler.jsonc
{
  "name": "frontend",
  "placement": { "mode": "off" },  // Explicit: stay at edge
  "services": [{ "binding": "BACKEND", "service": "backend-api" }]
}

// backend/wrangler.jsonc
{
  "name": "backend-api",
  "placement": { "mode": "smart" },
  "d1_databases": [{ "binding": "DB", "database_id": "xxx" }]
}
```

## Local Development Confusion

**Issue:** Smart Placement doesn't work in `wrangler dev`.

**Explanation:** Smart Placement only activates in production deployments, not local development.

**Solution:** Test Smart Placement in staging environment: `wrangler deploy --env staging`

## Baseline Traffic & Analysis Time

**Note:** Smart Placement routes 1% of requests WITHOUT optimization for comparison (expected).

**Analysis time:** Up to 15 minutes. During analysis, Worker runs at edge. Monitor `placement_status`.

## RPC Methods Not Affected (Critical Limitation)

**Problem:** Enabled Smart Placement on backend but RPC calls still slow.

**Cause:** Smart Placement ONLY affects `fetch` handlers. RPC methods (Service Bindings with `WorkerEntrypoint`) are NEVER affected.

**Why:** RPC bypasses `fetch` handler - Smart Placement can only route `fetch` requests.

**Solution:** Convert to fetch-based Service Bindings:

```typescript
// ❌ RPC - Smart Placement has NO EFFECT
export class BackendRPC extends WorkerEntrypoint {
  async getData() {
    // ALWAYS runs at edge
    return await this.env.DATABASE.prepare('SELECT * FROM table').all();
  }
}

// ✅ Fetch - Smart Placement WORKS
export default {
  async fetch(request: Request, env: Env): Promise<Response> {
    // Runs close to DATABASE when Smart Placement enabled
    const data = await env.DATABASE.prepare('SELECT * FROM table').all();
    return Response.json(data);
  }
}
```

## Requirements

- **Wrangler 2.20.0+** required
- **Consistent multi-region traffic** needed for analysis
- **Only affects fetch handlers** - RPC methods and named entrypoints not affected

## Limits

| Resource/Limit | Value | Notes |
|----------------|-------|-------|
| Analysis time | Up to 15 minutes | After enabling |
| Baseline traffic | 1% | Routed without optimization |
| Min Wrangler version | 2.20.0+ | Required |
| Traffic requirement | Multi-region | Consistent needed |

## Disabling Smart Placement

```jsonc
{ "placement": { "mode": "off" } }  // Explicit disable
// OR remove "placement" field entirely (same effect)
```

Both behaviors identical - Worker runs at edge closest to user.

## When NOT to Use Smart Placement

- Workers serving only static content or cached responses
- Workers without significant backend communication
- Pure edge logic (auth checks, redirects, simple transformations)
- Workers without fetch event handlers
- Pages/Assets Workers with `run_worker_first = true`
- Workers using RPC methods instead of fetch handlers

These scenarios won't benefit and may perform worse with Smart Placement.

# Smart Placement Patterns

## Backend Worker with Database Access

```typescript
export default {
  async fetch(request: Request, env: Env): Promise<Response> {
    const user = await env.DATABASE.prepare('SELECT * FROM users WHERE id = ?').bind(userId).first();
    const orders = await env.DATABASE.prepare('SELECT * FROM orders WHERE user_id = ?').bind(userId).all();
    return Response.json({ user, orders });
  }
};
```

```jsonc
{ "placement": { "mode": "smart" }, "d1_databases": [{ "binding": "DATABASE", "database_id": "xxx" }] }
```

## Frontend + Backend Split (Service Bindings)

**Frontend:** Runs at edge for fast user response
**Backend:** Smart Placement runs close to database

```typescript
// Frontend Worker - routes requests to backend
interface Env {
  BACKEND: Fetcher;  // Service Binding to backend Worker
}

export default {
  async fetch(request: Request, env: Env): Promise<Response> {
    if (new URL(request.url).pathname.startsWith('/api/')) {
      return env.BACKEND.fetch(request);  // Forward to backend
    }
    return new Response('Frontend content');
  }
};

// Backend Worker - database operations
interface BackendEnv {
  DATABASE: D1Database;
}

export default {
  async fetch(request: Request, env: BackendEnv): Promise<Response> {
    const data = await env.DATABASE.prepare('SELECT * FROM table').all();
    return Response.json(data);
  }
};
```

**CRITICAL:** Use fetch-based Service Bindings (shown above). If using RPC with `WorkerEntrypoint`, Smart Placement will NOT optimize those method calls - only `fetch` handlers are affected.

**RPC vs Fetch - CRITICAL:** Smart Placement ONLY works with fetch-based bindings, NOT RPC.

```typescript
// ❌ RPC - Smart Placement has NO EFFECT on backend RPC methods
export class BackendRPC extends WorkerEntrypoint {
  async getData() {
    // ALWAYS runs at edge, Smart Placement ignored
    return await this.env.DATABASE.prepare('SELECT * FROM table').all();
  }
}

// ✅ Fetch - Smart Placement WORKS
export default {
  async fetch(request: Request, env: Env): Promise<Response> {
    // Runs close to DATABASE when Smart Placement enabled
    const data = await env.DATABASE.prepare('SELECT * FROM table').all();
    return Response.json(data);
  }
};
```

## External API Integration

```typescript
export default {
  async fetch(request: Request, env: Env): Promise<Response> {
    const apiUrl = 'https://api.partner.com';
    const headers = { 'Authorization': `Bearer ${env.API_KEY}` };
    
    const [profile, transactions] = await Promise.all([
      fetch(`${apiUrl}/profile`, { headers }),
      fetch(`${apiUrl}/transactions`, { headers })
    ]);
    
    return Response.json({ 
      profile: await profile.json(), 
      transactions: await transactions.json()
    });
  }
};
```

## SSR / API Gateway Pattern

```typescript
// Frontend (edge) - auth/routing close to user
export default {
  async fetch(request: Request, env: Env) {
    if (!request.headers.get('Authorization')) {
      return new Response('Unauthorized', { status: 401 });
    }
    const data = await env.BACKEND.fetch(request);
    return new Response(renderPage(await data.json()), { 
      headers: { 'Content-Type': 'text/html' } 
    });
  }
};

// Backend (Smart Placement) - DB operations close to data
export default {
  async fetch(request: Request, env: Env) {
    const data = await env.DATABASE.prepare('SELECT * FROM pages WHERE id = ?').bind(pageId).first();
    return Response.json(data);
  }
};
```

## Durable Objects with Smart Placement

**Key principle:** Smart Placement does NOT control WHERE Durable Objects run. DOs always run in their designated region (based on jurisdiction or smart location hints).

**What Smart Placement DOES affect:** The location of the coordinator Worker's `fetch` handler that makes calls to multiple DOs.

**Pattern:** Enable Smart Placement on coordinator Worker that aggregates data from multiple DOs:

```typescript
// Worker with Smart Placement - aggregates data from multiple DOs
export default {
  async fetch(request: Request, env: Env): Promise<Response> {
    const userId = new URL(request.url).searchParams.get('user');
    
    // Get DO stubs
    const userDO = env.USER_DO.get(env.USER_DO.idFromName(userId));
    const analyticsID = env.ANALYTICS_DO.idFromName(`analytics-${userId}`);
    const analyticsDO = env.ANALYTICS_DO.get(analyticsID);
    
    // Fetch from multiple DOs
    const [userData, analyticsData] = await Promise.all([
      userDO.fetch(new Request('https://do/profile')),
      analyticsDO.fetch(new Request('https://do/stats'))
    ]);
    
    return Response.json({
      user: await userData.json(),
      analytics: await analyticsData.json()
    });
  }
};
```

```jsonc
// wrangler.jsonc
{
  "placement": { "mode": "smart" },
  "durable_objects": {
    "bindings": [
      { "name": "USER_DO", "class_name": "UserDO" },
      { "name": "ANALYTICS_DO", "class_name": "AnalyticsDO" }
    ]
  }
}
```

**When this helps:** 
- Worker's `fetch` handler runs closer to DO regions, reducing network latency for multiple DO calls
- Most beneficial when DOs are geographically concentrated or in specific jurisdictions
- Helps when coordinator makes many sequential or parallel DO calls

**When this DOESN'T help:**
- DOs are globally distributed (no single optimal Worker location)
- Worker only calls a single DO
- DO calls are infrequent or cached

## Best Practices

- Split full-stack apps: frontend at edge, backend with Smart Placement
- Use fetch-based Service Bindings (not RPC)
- Enable for backend logic: APIs, data aggregation, DB operations
- Don't enable for: static content, edge logic, RPC methods, Pages with `run_worker_first`
- Wait 15+ min for analysis, verify `placement_status = SUCCESS`

# Cloudflare Workers Smart Placement

Automatic workload placement optimization to minimize latency by running Workers closer to backend infrastructure rather than end users.

## Core Concept

Smart Placement automatically analyzes Worker request duration across Cloudflare's global network and intelligently routes requests to optimal data center locations. Instead of defaulting to the location closest to the end user, Smart Placement can forward requests to locations closer to backend infrastructure when this reduces overall request duration.

### When to Use

**Enable Smart Placement when:**
- Worker makes multiple round trips to backend services/databases
- Backend infrastructure is geographically concentrated
- Request duration dominated by backend latency rather than network latency from user
- Running backend logic in Workers (APIs, data aggregation, SSR with DB calls)
- Worker uses `fetch` handler (not RPC methods)

**Do NOT enable for:**
- Workers serving only static content or cached responses
- Workers without significant backend communication
- Pure edge logic (auth checks, redirects, simple transformations)
- Workers without fetch event handlers
- Workers with RPC methods or named entrypoints (only `fetch` handlers are affected)
- Pages/Assets Workers with `run_worker_first = true` (degrades asset serving)

### Decision Tree

```
Does your Worker have a fetch handler?
├─ No → Smart Placement won't work (skip)
└─ Yes
   │
   Does it make multiple backend calls (DB/API)?
   ├─ No → Don't enable (won't help)
   └─ Yes
      │
      Is backend geographically concentrated?
      ├─ No (globally distributed) → Probably won't help
      └─ Yes or uncertain
         │
         Does it serve static assets with run_worker_first=true?
         ├─ Yes → Don't enable (will hurt performance)
         └─ No → Enable Smart Placement
            │
            After 15min, check placement_status
            ├─ SUCCESS → Monitor metrics
            ├─ INSUFFICIENT_INVOCATIONS → Need more traffic
            └─ UNSUPPORTED_APPLICATION → Disable (hurting performance)
```

### Key Architecture Pattern

**Recommended:** Split full-stack applications into separate Workers:
```
User → Frontend Worker (at edge, close to user)
         ↓ Service Binding
       Backend Worker (Smart Placement enabled, close to DB/API)
         ↓
       Database/Backend Service
```

This maintains fast, reactive frontends while optimizing backend latency.

## Quick Start

```jsonc
// wrangler.jsonc
{
  "placement": {
    "mode": "smart"  // or "off" to explicitly disable
  }
}
```

Deploy and wait 15 minutes for analysis. Check status via API or dashboard metrics.

**To disable:** Set `"mode": "off"` or remove `placement` field entirely (both equivalent).

## Requirements

- Wrangler 2.20.0+
- Analysis time: Up to 15 minutes after enabling
- Traffic requirements: Consistent traffic from multiple global locations
- Available on all Workers plans (Free, Paid, Enterprise)

## Placement Status Values

```typescript
type PlacementStatus = 
  | undefined  // Not yet analyzed
  | 'SUCCESS'  // Successfully optimized
  | 'INSUFFICIENT_INVOCATIONS'  // Not enough traffic
  | 'UNSUPPORTED_APPLICATION';  // Made Worker slower (reverted)
```

## CLI Commands

```bash
# Deploy with Smart Placement
wrangler deploy

# Check placement status
curl -H "Authorization: Bearer $TOKEN" \
  https://api.cloudflare.com/client/v4/accounts/$ACCOUNT_ID/workers/services/$WORKER_NAME \
  | jq .result.placement_status

# Monitor
wrangler tail your-worker-name --header cf-placement
```

## Reading Order

**First time?** Start here:
1. This README - understand core concepts and when to use Smart Placement
2. [configuration.md](./configuration.md) - set up wrangler.jsonc and understand limitations
3. [patterns.md](./patterns.md) - see practical examples for your use case
4. [api.md](./api.md) - monitor and verify Smart Placement is working
5. [gotchas.md](./gotchas.md) - troubleshoot common issues

**Quick lookup:**
- "Should I enable Smart Placement?" → See "When to Use" above
- "How do I configure it?" → [configuration.md](./configuration.md)
- "How do I split frontend/backend?" → [patterns.md](./patterns.md)
- "Why isn't it working?" → [gotchas.md](./gotchas.md)

## In This Reference

- [configuration.md](./configuration.md) - wrangler.jsonc setup, mode values, validation rules
- [api.md](./api.md) - Placement Status API, cf-placement header, monitoring
- [patterns.md](./patterns.md) - Frontend/backend split, database workers, SSR patterns
- [gotchas.md](./gotchas.md) - Troubleshooting INSUFFICIENT_INVOCATIONS, performance issues

## See Also

- [workers](https://developers.cloudflare.com/workers/) - Worker runtime and fetch handlers
- [d1](../d1/) - D1 database that benefits from Smart Placement
- [durable-objects](https://developers.cloudflare.com/durable-objects/) - Durable Objects with backend logic
- [bindings](../bindings/) - Service bindings for frontend/backend split

# Snippets API Reference

## Request Object

### HTTP Properties
```javascript
request.method    // GET, POST, PUT, DELETE, etc.
request.url       // Full URL string
request.headers   // Headers object
request.body      // ReadableStream (for POST/PUT)
request.cf        // Cloudflare properties (see below)
```

### URL Operations
```javascript
const url = new URL(request.url);
url.hostname             // "example.com"
url.pathname             // "/path/to/page"
url.search               // "?query=value"
url.searchParams.get("q") // "value"
url.searchParams.set("q", "new")
url.searchParams.delete("q")
```

### Header Operations
```javascript
// Read headers
request.headers.get("User-Agent")
request.headers.has("Authorization")
request.headers.getSetCookie() // Get all Set-Cookie headers

// Modify headers (create new request)
const modifiedRequest = new Request(request);
modifiedRequest.headers.set("X-Custom", "value")
modifiedRequest.headers.delete("X-Remove")
```

### Cloudflare Properties (`request.cf`)
Access Cloudflare-specific metadata about the request:

```javascript
// Geolocation
request.cf.city            // "San Francisco"
request.cf.continent       // "NA"
request.cf.country         // "US"
request.cf.region          // "California" or "CA"
request.cf.regionCode      // "CA"
request.cf.postalCode      // "94102"
request.cf.latitude        // "37.7749"
request.cf.longitude       // "-122.4194"
request.cf.timezone        // "America/Los_Angeles"
request.cf.metroCode       // "807" (DMA code)

// Network
request.cf.colo            // "SFO" (airport code of datacenter)
request.cf.asn             // 13335 (ASN number)
request.cf.asOrganization  // "Cloudflare, Inc."

// Bot Management (if enabled)
request.cf.botManagement.score        // 1-99 (1=bot, 99=human)
request.cf.botManagement.verified_bot // true/false
request.cf.botManagement.static_resource // true/false

// TLS/HTTP version
request.cf.tlsVersion      // "TLSv1.3"
request.cf.tlsCipher       // "AEAD-AES128-GCM-SHA256"
request.cf.httpProtocol    // "HTTP/2"

// Request metadata
request.cf.requestPriority // "weight=192;exclusive=0"
```

**Use cases**: Geo-routing, bot detection, security decisions, analytics.

## Response Object

### Response Constructors
```javascript
// Plain text
new Response("Hello", { status: 200 })

// JSON
Response.json({ key: "value" }, { status: 200 })

// HTML
new Response("<h1>Hi</h1>", { 
  status: 200,
  headers: { "Content-Type": "text/html" }
})

// Redirect
Response.redirect("https://example.com", 301) // or 302

// Stream (pass through)
new Response(response.body, response)
```

### Response Headers
```javascript
// Create modified response
const newResponse = new Response(response.body, response);

// Set/modify headers
newResponse.headers.set("X-Custom", "value")
newResponse.headers.append("Set-Cookie", "session=abc; Path=/")
newResponse.headers.delete("Server")

// Common headers
newResponse.headers.set("Cache-Control", "public, max-age=3600")
newResponse.headers.set("Content-Type", "application/json")
```

### Response Properties
```javascript
response.status       // 200, 404, 500, etc.
response.statusText   // "OK", "Not Found", etc.
response.headers      // Headers object
response.body         // ReadableStream
response.ok           // true if status 200-299
response.redirected   // true if redirected
```

## REST API Operations

### List Snippets
```bash
GET /zones/{zone_id}/snippets
```

### Get Snippet
```bash
GET /zones/{zone_id}/snippets/{snippet_name}
```

### Create/Update Snippet
```bash
PUT /zones/{zone_id}/snippets/{snippet_name}
Content-Type: multipart/form-data

files=@snippet.js
metadata={"main_module":"snippet.js"}
```

### Delete Snippet
```bash
DELETE /zones/{zone_id}/snippets/{snippet_name}
```

### List Snippet Rules
```bash
GET /zones/{zone_id}/rulesets/phases/http_request_snippets/entrypoint
```

### Update Snippet Rules
```bash
PUT /zones/{zone_id}/snippets/snippet_rules
Content-Type: application/json

{
  "rules": [{
    "description": "Apply snippet",
    "enabled": true,
    "expression": "http.host eq \"example.com\"",
    "snippet_name": "my_snippet"
  }]
}
```

## Available APIs in Snippets

### ✅ Supported
- `fetch()` - HTTP requests (2-5 subrequests per plan)
- `Request` / `Response` - Standard Web APIs
- `URL` / `URLSearchParams` - URL manipulation
- `Headers` - Header manipulation
- `TextEncoder` / `TextDecoder` - Text encoding
- `crypto.subtle` - Web Crypto API (hashing, signing)
- `crypto.randomUUID()` - UUID generation

### ❌ Not Supported in Snippets
- `caches` API - Not available (use Workers)
- `KV`, `D1`, `R2` - Storage APIs (use Workers)
- `Durable Objects` - Stateful objects (use Workers)
- `WebSocket` - WebSocket upgrades (use Workers)
- `HTMLRewriter` - HTML parsing (use Workers)
- `import` statements - No module imports
- `addEventListener` - Use `export default { async fetch() {}` pattern

## Snippet Structure
```javascript
export default {
  async fetch(request) {
    // Your logic here
    const response = await fetch(request);
    return response; // or modified response
  }
}
```

# Snippets Configuration Guide

## Configuration Methods

### 1. Dashboard (GUI)
**Best for**: Quick tests, single snippets, visual rule building

```
1. Go to zone → Rules → Snippets
2. Click "Create Snippet" or select template
3. Enter snippet name (a-z, 0-9, _ only, cannot change later)
4. Write JavaScript code (32KB max)
5. Configure snippet rule:
   - Expression Builder (visual) or Expression Editor (text)
   - Use Ruleset Engine filter expressions
6. Test with Preview/HTTP tabs
7. Deploy or Save as Draft
```

### 2. REST API
**Best for**: CI/CD, automation, programmatic management

```bash
# Create/update snippet
curl "https://api.cloudflare.com/client/v4/zones/$ZONE_ID/snippets/$SNIPPET_NAME" \
  --request PUT \
  --header "Authorization: Bearer $CLOUDFLARE_API_TOKEN" \
  --form "files=@example.js" \
  --form "metadata={\"main_module\": \"example.js\"}"

# Create snippet rule
curl "https://api.cloudflare.com/client/v4/zones/$ZONE_ID/snippets/snippet_rules" \
  --request PUT \
  --header "Authorization: Bearer $CLOUDFLARE_API_TOKEN" \
  --header "Content-Type: application/json" \
  --data '{
    "rules": [
      {
        "description": "Trigger snippet on /api paths",
        "enabled": true,
        "expression": "starts_with(http.request.uri.path, \"/api/\")",
        "snippet_name": "api_snippet"
      }
    ]
  }'

# List snippets
curl "https://api.cloudflare.com/client/v4/zones/$ZONE_ID/snippets" \
  --header "Authorization: Bearer $CLOUDFLARE_API_TOKEN"

# Delete snippet
curl "https://api.cloudflare.com/client/v4/zones/$ZONE_ID/snippets/$SNIPPET_NAME" \
  --request DELETE \
  --header "Authorization: Bearer $CLOUDFLARE_API_TOKEN"
```

### 3. Terraform
**Best for**: Infrastructure-as-code, multi-zone deployments

```hcl
# Configure Terraform provider
terraform {
  required_providers {
    cloudflare = {
      source  = "cloudflare/cloudflare"
      version = "~> 4.0"
    }
  }
}

provider "cloudflare" {
  api_token = var.cloudflare_api_token
}

# Create snippet
resource "cloudflare_snippet" "security_headers" {
  zone_id = var.zone_id
  name    = "security_headers"
  
  main_module = "security_headers.js"
  files {
    name    = "security_headers.js"
    content = file("${path.module}/snippets/security_headers.js")
  }
}

# Create snippet rule
resource "cloudflare_snippet_rules" "security_rules" {
  zone_id = var.zone_id
  
  rules {
    description  = "Apply security headers to all requests"
    enabled      = true
    expression   = "true"
    snippet_name = cloudflare_snippet.security_headers.name
  }
}
```

### 4. Pulumi
**Best for**: Multi-cloud IaC, TypeScript/Python/Go workflows

```typescript
import * as cloudflare from "@pulumi/cloudflare";
import * as fs from "fs";

// Create snippet
const securitySnippet = new cloudflare.Snippet("security-headers", {
  zoneId: zoneId,
  name: "security_headers",
  mainModule: "security_headers.js",
  files: [{
    name: "security_headers.js",
    content: fs.readFileSync("./snippets/security_headers.js", "utf8"),
  }],
});

// Create snippet rule
const snippetRule = new cloudflare.SnippetRules("security-rules", {
  zoneId: zoneId,
  rules: [{
    description: "Apply security headers",
    enabled: true,
    expression: "true",
    snippetName: securitySnippet.name,
  }],
});
```

## Filter Expressions

Snippets use Cloudflare's Ruleset Engine expression language to determine when to execute.

### Common Expression Patterns

```javascript
// Host matching
http.host eq "example.com"
http.host in {"example.com" "www.example.com"}
http.host contains "example"

// Path matching
http.request.uri.path eq "/api/users"
starts_with(http.request.uri.path, "/api/")
ends_with(http.request.uri.path, ".json")
matches(http.request.uri.path, "^/api/v[0-9]+/")

// Query parameters
http.request.uri.query contains "debug=true"

// Headers
http.headers["user-agent"] contains "Mobile"
http.headers["accept-language"] eq "en-US"

// Cookies
http.cookie contains "session="

// Geolocation
ip.geoip.country eq "US"
ip.geoip.continent eq "EU"

// Bot detection (requires Bot Management)
cf.bot_management.score lt 30

// Method
http.request.method eq "POST"
http.request.method in {"POST" "PUT" "PATCH"}

// Combine with logical operators
http.host eq "example.com" and starts_with(http.request.uri.path, "/api/")
ip.geoip.country eq "US" or ip.geoip.country eq "CA"
not http.headers["user-agent"] contains "bot"
```

### Expression Functions

| Function | Example | Description |
|----------|---------|-------------|
| `starts_with()` | `starts_with(http.request.uri.path, "/api/")` | Check prefix |
| `ends_with()` | `ends_with(http.request.uri.path, ".json")` | Check suffix |
| `contains()` | `contains(http.headers["user-agent"], "Mobile")` | Check substring |
| `matches()` | `matches(http.request.uri.path, "^/api/")` | Regex match |
| `lower()` | `lower(http.host) eq "example.com"` | Convert to lowercase |
| `upper()` | `upper(http.headers["x-api-key"])` | Convert to uppercase |
| `len()` | `len(http.request.uri.path) gt 100` | String length |

## Deployment Workflow

### Development
1. Write snippet code locally
2. Test syntax with `node snippet.js` or TypeScript compiler
3. Deploy to Dashboard or use API with `Save as Draft`
4. Test with Preview/HTTP tabs in Dashboard
5. Enable rule when ready

### Production
1. Store snippet code in version control
2. Use Terraform/Pulumi for reproducible deployments
3. Deploy to staging zone first
4. Test with real traffic (use low-traffic subdomain)
5. Apply to production zone
6. Monitor with Analytics/Logpush

## Limits & Requirements

| Resource | Limit | Notes |
|----------|-------|-------|
| Snippet size | 32 KB | Per snippet, compressed |
| Snippet name | 64 chars | `a-z`, `0-9`, `_` only, immutable |
| Snippets per zone | 20 | Soft limit, contact support for more |
| Rules per zone | 20 | One rule per snippet typical |
| Expression length | 4096 chars | Per rule expression |

## Authentication

### API Token (Recommended)
```bash
# Create token at: https://dash.cloudflare.com/profile/api-tokens
# Required permissions: Zone.Snippets:Edit, Zone.Rules:Edit
export CLOUDFLARE_API_TOKEN="your_token_here"
```

### API Key (Legacy)
```bash
export CLOUDFLARE_EMAIL="your@email.com"
export CLOUDFLARE_API_KEY="your_global_api_key"
``` 

# Gotchas & Best Practices

## Common Errors

### 1000: "Snippet execution failed"
Runtime error or syntax error. Wrap code in try/catch:
```javascript
try { return await fetch(request); }
catch (error) { return new Response(`Error: ${error.message}`, { status: 500 }); }
```

### 1100: "Exceeded execution limit"
Code takes >5ms CPU. Simplify logic or move to Workers.

### 1201: "Multiple origin fetches"
Call `fetch(request)` exactly once:
```javascript
// ❌ Multiple origin fetches
const r1 = await fetch(request); const r2 = await fetch(request);
// ✅ Single fetch, reuse response
const response = await fetch(request);
```

### 1202: "Subrequest limit exceeded"
Pro: 2 subrequests, Business/Enterprise: 5. Reduce fetch calls.

### "Cannot set property on immutable object"
Clone before modifying:
```javascript
const modifiedRequest = new Request(request);
modifiedRequest.headers.set("X-Custom", "value");
```

### "caches is not defined"
Cache API NOT available in Snippets. Use Workers.

### "Module not found"
Snippets don't support `import`. Use inline code or Workers.

## Best Practices

### Performance
- Keep code <10KB (32KB limit)
- Optimize for 5ms CPU
- Clone only when modifying
- Minimize subrequests

### Security
- Validate all inputs
- Use Web Crypto API for hashing
- Sanitize headers before origin
- Don't log secrets

### Debugging
```javascript
newResponse.headers.set("X-Debug-Country", request.cf.country);
```
```bash
curl -H "X-Test: true" https://example.com -v
```

## Available APIs

**✅ Available:** `fetch()`, `Request`, `Response`, `Headers`, `URL`, `crypto.subtle`, `crypto.randomUUID()`, `atob()`/`btoa()`, `JSON`

**❌ NOT Available:** `caches`, `KV`, `D1`, `R2`, `Durable Objects`, `WebSocket`, `HTMLRewriter`, `import`, Node.js APIs

## Limits

| Resource | Limit |
|----------|-------|
| Snippet size | 32KB |
| Execution time | 5ms CPU |
| Subrequests (Pro/Biz) | 2/5 |
| Snippets/zone | 20 |

## Performance Benchmarks

| Operation | Time |
|-----------|------|
| Header set | <0.1ms |
| URL parsing | <0.2ms |
| fetch() | 1-3ms |
| SHA-256 | 0.5-1ms |

**Migrate to Workers when:** >5ms needed, >5 subrequests, need storage (KV/D1/R2), need npm packages, >32KB code

# Snippets Patterns

## Security Headers

```javascript
export default {
  async fetch(request) {
    const response = await fetch(request);
    const newResponse = new Response(response.body, response);
    newResponse.headers.set("X-Frame-Options", "DENY");
    newResponse.headers.set("X-Content-Type-Options", "nosniff");
    newResponse.headers.delete("X-Powered-By");
    return newResponse;
  }
}
```

**Rule:** `true` (all requests)

## Geo-Based Routing

```javascript
export default {
  async fetch(request) {
    const country = request.cf.country;
    if (["GB", "DE", "FR"].includes(country)) {
      const url = new URL(request.url);
      url.hostname = url.hostname.replace(".com", ".eu");
      return Response.redirect(url.toString(), 302);
    }
    return fetch(request);
  }
}
```

## A/B Testing

```javascript
export default {
  async fetch(request) {
    const cookies = request.headers.get("Cookie") || "";
    let variant = cookies.match(/ab_test=([AB])/)?.[1] || (Math.random() < 0.5 ? "A" : "B");
    
    const req = new Request(request);
    req.headers.set("X-Variant", variant);
    const response = await fetch(req);
    
    if (!cookies.includes("ab_test=")) {
      const newResponse = new Response(response.body, response);
      newResponse.headers.append("Set-Cookie", `ab_test=${variant}; Path=/; Secure`);
      return newResponse;
    }
    return response;
  }
}
```

## Bot Detection

```javascript
export default {
  async fetch(request) {
    const botScore = request.cf.botManagement?.score;
    if (botScore && botScore < 30) return new Response("Denied", { status: 403 });
    return fetch(request);
  }
}
```

**Requires:** Bot Management plan

## API Auth Header Injection

```javascript
export default {
  async fetch(request) {
    if (new URL(request.url).pathname.startsWith("/api/")) {
      const req = new Request(request);
      req.headers.set("X-Internal-Auth", "secret_token");
      req.headers.delete("Authorization");
      return fetch(req);
    }
    return fetch(request);
  }
}
```

## CORS Headers

```javascript
export default {
  async fetch(request) {
    if (request.method === "OPTIONS") {
      return new Response(null, {
        status: 204,
        headers: {
          "Access-Control-Allow-Origin": "*",
          "Access-Control-Allow-Methods": "GET, POST, PUT, DELETE",
          "Access-Control-Allow-Headers": "Content-Type, Authorization"
        }
      });
    }
    const response = await fetch(request);
    const newResponse = new Response(response.body, response);
    newResponse.headers.set("Access-Control-Allow-Origin", "*");
    return newResponse;
  }
}
```

## Maintenance Mode

```javascript
export default {
  async fetch(request) {
    if (request.headers.get("X-Bypass-Token") === "admin") return fetch(request);
    return new Response("<h1>Maintenance</h1>", {
      status: 503,
      headers: { "Content-Type": "text/html", "Retry-After": "3600" }
    });
  }
}
```

## Pattern Selection

| Pattern | Complexity | Use Case |
|---------|-----------|----------|
| Security Headers | Low | All sites |
| Geo-Routing | Low | Regional content |
| A/B Testing | Medium | Experiments |
| Bot Detection | Medium | Requires Bot Management |
| API Auth | Low | Backend protection |
| CORS | Low | API endpoints |
| Maintenance | Low | Deployments |

# Cloudflare Snippets Skill Reference

## Description
Expert guidance for **Cloudflare Snippets ONLY** - a lightweight JavaScript-based edge logic platform for modifying HTTP requests and responses. Snippets run as part of the Ruleset Engine and are included at no additional cost on paid plans (Pro, Business, Enterprise).

## What Are Snippets?
Snippets are JavaScript functions executed at the edge as part of Cloudflare's Ruleset Engine. Key characteristics:
- **Execution time**: 5ms CPU limit per request
- **Size limit**: 32KB per snippet
- **Runtime**: V8 isolate (subset of Workers APIs)
- **Subrequests**: 2-5 fetch calls depending on plan
- **Cost**: Included with Pro/Business/Enterprise plans

## Snippets vs Workers Decision Matrix

| Factor | Choose Snippets If... | Choose Workers If... |
|--------|----------------------|---------------------|
| **Complexity** | Simple request/response modifications | Complex business logic, routing, middleware |
| **Execution time** | <5ms sufficient | Need >5ms or variable time |
| **Subrequests** | 2-5 fetch calls sufficient | Need >5 subrequests or complex orchestration |
| **Code size** | <32KB sufficient | Need >32KB or npm dependencies |
| **Cost** | Want zero additional cost | Can afford $5/mo + usage |
| **APIs** | Need basic fetch, headers, URL | Need KV, D1, R2, Durable Objects, cron triggers |
| **Deployment** | Need rule-based triggers | Want custom routing logic |

**Rule of thumb**: Use Snippets for modifications, Workers for applications.

## Execution Model
1. Request arrives at Cloudflare edge
2. Ruleset Engine evaluates snippet rules (filter expressions)
3. If rule matches, snippet executes within 5ms limit
4. Modified request/response continues through pipeline
5. Response returned to client

Snippets execute synchronously in the request path - performance is critical.

## Reading Order
1. **[configuration.md](configuration.md)** - Start here: setup, deployment methods (Dashboard/API/Terraform)
2. **[api.md](api.md)** - Core APIs: Request, Response, headers, `request.cf` properties
3. **[patterns.md](patterns.md)** - Real-world examples: geo-routing, A/B tests, security headers
4. **[gotchas.md](gotchas.md)** - Troubleshooting: common errors, performance tips, API limitations

## In This Reference

- **[configuration.md](configuration.md)** - Setup, deployment, configuration
- **[api.md](api.md)** - API endpoints, methods, interfaces
- **[patterns.md](patterns.md)** - Common patterns, use cases, examples
- **[gotchas.md](gotchas.md)** - Troubleshooting, best practices, limitations

## Quick Start
```javascript
// Snippet: Add security headers
export default {
  async fetch(request) {
    const response = await fetch(request);
    const newResponse = new Response(response.body, response);
    newResponse.headers.set("X-Frame-Options", "DENY");
    newResponse.headers.set("X-Content-Type-Options", "nosniff");
    return newResponse;
  }
}
```

Deploy via Dashboard (Rules → Snippets) or API/Terraform. See configuration.md for details.

## See Also

- [Cloudflare Docs](https://developers.cloudflare.com/rules/snippets/)

## REST API Endpoints

```
GET    /zones/{zone_id}/spectrum/apps                    # List apps
POST   /zones/{zone_id}/spectrum/apps                    # Create app
GET    /zones/{zone_id}/spectrum/apps/{app_id}           # Get app
PUT    /zones/{zone_id}/spectrum/apps/{app_id}           # Update app
DELETE /zones/{zone_id}/spectrum/apps/{app_id}           # Delete app

GET    /zones/{zone_id}/spectrum/analytics/aggregate/current
GET    /zones/{zone_id}/spectrum/analytics/events/bytime
GET    /zones/{zone_id}/spectrum/analytics/events/summary
```

## Request/Response Schemas

### CreateSpectrumAppRequest

```typescript
interface CreateSpectrumAppRequest {
  protocol: string;                    // "tcp/22", "udp/53"
  dns: {
    type: "CNAME" | "ADDRESS";
    name: string;                      // "ssh.example.com"
  };
  origin_direct?: string[];            // ["tcp://192.0.2.1:22"]
  origin_dns?: { name: string };       // {"name": "origin.example.com"}
  origin_port?: number | { start: number; end: number };
  proxy_protocol?: "off" | "v1" | "v2" | "simple";
  ip_firewall?: boolean;
  tls?: "off" | "flexible" | "full" | "strict";
  edge_ips?: {
    type: "dynamic" | "static";
    connectivity: "all" | "ipv4" | "ipv6";
  };
  traffic_type?: "direct" | "http" | "https";
  argo_smart_routing?: boolean;
}
```

### SpectrumApp Response

```typescript
interface SpectrumApp {
  id: string;
  protocol: string;
  dns: { type: string; name: string };
  origin_direct?: string[];
  origin_dns?: { name: string };
  origin_port?: number | { start: number; end: number };
  proxy_protocol: string;
  ip_firewall: boolean;
  tls: string;
  edge_ips: { type: string; connectivity: string; ips?: string[] };
  argo_smart_routing: boolean;
  created_on: string;
  modified_on: string;
}
```

## TypeScript SDK

```typescript
import Cloudflare from 'cloudflare';

const client = new Cloudflare({ apiToken: process.env.CLOUDFLARE_API_TOKEN });

// Create
const app = await client.spectrum.apps.create({
  zone_id: 'your-zone-id',
  protocol: 'tcp/22',
  dns: { type: 'CNAME', name: 'ssh.example.com' },
  origin_direct: ['tcp://192.0.2.1:22'],
  ip_firewall: true,
  tls: 'off',
});

// List
const apps = await client.spectrum.apps.list({ zone_id: 'your-zone-id' });

// Get
const appDetails = await client.spectrum.apps.get({ zone_id: 'your-zone-id', app_id: app.id });

// Update
await client.spectrum.apps.update({ zone_id: 'your-zone-id', app_id: app.id, tls: 'full' });

// Delete
await client.spectrum.apps.delete({ zone_id: 'your-zone-id', app_id: app.id });

// Analytics
const analytics = await client.spectrum.analytics.aggregate({
  zone_id: 'your-zone-id',
  metrics: ['bytesIngress', 'bytesEgress'],
  since: new Date(Date.now() - 3600000).toISOString(),
});
```

## Python SDK

```python
from cloudflare import Cloudflare

client = Cloudflare(api_token="your-api-token")

# Create
app = client.spectrum.apps.create(
    zone_id="your-zone-id",
    protocol="tcp/22",
    dns={"type": "CNAME", "name": "ssh.example.com"},
    origin_direct=["tcp://192.0.2.1:22"],
    ip_firewall=True,
    tls="off",
)

# List
apps = client.spectrum.apps.list(zone_id="your-zone-id")

# Get
app_details = client.spectrum.apps.get(zone_id="your-zone-id", app_id=app.id)

# Update
client.spectrum.apps.update(zone_id="your-zone-id", app_id=app.id, tls="full")

# Delete
client.spectrum.apps.delete(zone_id="your-zone-id", app_id=app.id)

# Analytics
analytics = client.spectrum.analytics.aggregate(
    zone_id="your-zone-id",
    metrics=["bytesIngress", "bytesEgress"],
    since=datetime.now() - timedelta(hours=1),
)
```

## Go SDK

```go
import "github.com/cloudflare/cloudflare-go"

api, _ := cloudflare.NewWithAPIToken("your-api-token")

// Create
app, _ := api.CreateSpectrumApplication(ctx, "zone-id", cloudflare.SpectrumApplication{
    Protocol:         "tcp/22",
    DNS:              cloudflare.SpectrumApplicationDNS{Type: "CNAME", Name: "ssh.example.com"},
    OriginDirect:     []string{"tcp://192.0.2.1:22"},
    IPFirewall:       true,
    ArgoSmartRouting: true,
})

// List
apps, _ := api.SpectrumApplications(ctx, "zone-id")

// Delete
_ = api.DeleteSpectrumApplication(ctx, "zone-id", app.ID)
```

## Analytics API

**Metrics:**
- `bytesIngress` - Bytes received from clients
- `bytesEgress` - Bytes sent to clients
- `count` - Number of connections
- `duration` - Connection duration (seconds)

**Dimensions:**
- `event` - Connection event type
- `appID` - Spectrum application ID
- `coloName` - Datacenter name
- `ipVersion` - IPv4 or IPv6

**Example:**
```bash
curl "https://api.cloudflare.com/client/v4/zones/$ZONE_ID/spectrum/analytics/aggregate/current?metrics=bytesIngress,bytesEgress,count&dimensions=appID" \
  --header "Authorization: Bearer $CLOUDFLARE_API_TOKEN"
```

## See Also

- [configuration.md](configuration.md) - Terraform/Pulumi
- [patterns.md](patterns.md) - Protocol examples

## Origin Types

### Direct IP Origin

Use when origin is a single server with static IP.

**TypeScript SDK:**
```typescript
const app = await client.spectrum.apps.create({
  zone_id: 'your-zone-id',
  protocol: 'tcp/22',
  dns: { type: 'CNAME', name: 'ssh.example.com' },
  origin_direct: ['tcp://192.0.2.1:22'],
  ip_firewall: true,
  tls: 'off',
});
```

**Terraform:**
```hcl
resource "cloudflare_spectrum_application" "ssh" {
  zone_id  = var.zone_id
  protocol = "tcp/22"

  dns {
    type = "CNAME"
    name = "ssh.example.com"
  }

  origin_direct      = ["tcp://192.0.2.1:22"]
  ip_firewall        = true
  tls                = "off"
  argo_smart_routing = true
}
```

### CNAME Origin

Use when origin is a hostname (not static IP). Spectrum resolves DNS dynamically.

**TypeScript SDK:**
```typescript
const app = await client.spectrum.apps.create({
  zone_id: 'your-zone-id',
  protocol: 'tcp/3306',
  dns: { type: 'CNAME', name: 'db.example.com' },
  origin_dns: { name: 'db-primary.internal.example.com' },
  origin_port: 3306,
  tls: 'full',
});
```

**Terraform:**
```hcl
resource "cloudflare_spectrum_application" "database" {
  zone_id  = var.zone_id
  protocol = "tcp/3306"

  dns {
    type = "CNAME"
    name = "db.example.com"
  }

  origin_dns {
    name = "db-primary.internal.example.com"
  }

  origin_port        = 3306
  tls                = "full"
  argo_smart_routing = true
}
```

### Load Balancer Origin

Use for high availability and failover.

**Terraform:**
```hcl
resource "cloudflare_load_balancer" "game_lb" {
  zone_id          = var.zone_id
  name             = "game-lb.example.com"
  default_pool_ids = [cloudflare_load_balancer_pool.game_pool.id]
}

resource "cloudflare_load_balancer_pool" "game_pool" {
  name    = "game-primary"
  origins { name = "game-1"; address = "192.0.2.1" }
  monitor = cloudflare_load_balancer_monitor.tcp_monitor.id
}

resource "cloudflare_load_balancer_monitor" "tcp_monitor" {
  type = "tcp"; port = 25565; interval = 60; timeout = 5
}

resource "cloudflare_spectrum_application" "game" {
  zone_id  = var.zone_id
  protocol = "tcp/25565"
  dns { type = "CNAME"; name = "game.example.com" }
  origin_dns { name = cloudflare_load_balancer.game_lb.name }
  origin_port = 25565
}
```

## TLS Configuration

| Mode | Description | Use Case | Origin Cert |
|------|-------------|----------|-------------|
| `off` | No TLS | Non-encrypted (SSH, gaming) | No |
| `flexible` | TLS client→CF, plain CF→origin | Testing | No |
| `full` | TLS end-to-end, self-signed OK | Production | Yes (any) |
| `strict` | Full + valid cert verification | Max security | Yes (CA) |

**Example:**
```typescript
const app = await client.spectrum.apps.create({
  zone_id: 'your-zone-id',
  protocol: 'tcp/3306',
  dns: { type: 'CNAME', name: 'db.example.com' },
  origin_direct: ['tcp://192.0.2.1:3306'],
  tls: 'strict',  // Validates origin certificate
});
```

## Proxy Protocol

Forwards real client IP to origin. Origin must support parsing.

| Version | Protocol | Use Case |
|---------|----------|----------|
| `off` | - | Origin doesn't need client IP |
| `v1` | TCP | Most TCP apps (SSH, databases) |
| `v2` | TCP | High-performance TCP |
| `simple` | UDP | UDP applications |

**Compatibility:**
- **v1**: HAProxy, nginx, SSH, most databases
- **v2**: HAProxy 1.5+, nginx 1.11+
- **simple**: Cloudflare-specific UDP format

**Enable:**
```typescript
const app = await client.spectrum.apps.create({
  // ...
  proxy_protocol: 'v1',  // Origin must parse PROXY header
});
```

**Origin Config (nginx):**
```nginx
stream {
    server {
        listen 22 proxy_protocol;
        proxy_pass backend:22;
    }
}
```

## IP Access Rules

Enable `ip_firewall: true` then configure zone-level firewall rules.

```typescript
const app = await client.spectrum.apps.create({
  // ...
  ip_firewall: true,  // Applies zone firewall rules
});
```

## Port Ranges (Enterprise Only)

```hcl
resource "cloudflare_spectrum_application" "game_cluster" {
  zone_id  = var.zone_id
  protocol = "tcp/25565-25575"

  dns {
    type = "CNAME"
    name = "games.example.com"
  }

  origin_direct = ["tcp://192.0.2.1"]
  
  origin_port {
    start = 25565
    end   = 25575
  }
}
```

## See Also

- [patterns.md](patterns.md) - Protocol-specific examples
- [api.md](api.md) - REST/SDK reference

## Common Issues

### Connection Timeouts

**Problem:** Connections fail or timeout  
**Cause:** Origin firewall blocking Cloudflare IPs, origin service not running, incorrect DNS  
**Solution:**
1. Verify origin firewall allows Cloudflare IP ranges
2. Check origin service running on correct port
3. Ensure DNS record is CNAME (not A/AAAA)
4. Verify origin IP/hostname is correct

```bash
# Test connectivity
nc -zv app.example.com 22
dig app.example.com
```

### Client IP Showing Cloudflare IP

**Problem:** Origin logs show Cloudflare IPs not real client IPs  
**Cause:** Proxy Protocol not enabled or origin not configured  
**Solution:**
```typescript
// Enable in Spectrum app
const app = await client.spectrum.apps.create({
  // ...
  proxy_protocol: 'v1',  // TCP: v1/v2; UDP: simple
});
```

**Origin config:**
- **nginx**: `listen 22 proxy_protocol;`
- **HAProxy**: `bind :22 accept-proxy`

### TLS Errors

**Problem:** TLS handshake failures, 525 errors  
**Cause:** TLS mode mismatch

| Error | TLS Mode | Problem | Solution |
|-------|----------|---------|----------|
| Connection refused | `full`/`strict` | Origin not TLS | Use `tls: "off"` or enable TLS |
| 525 cert invalid | `strict` | Self-signed cert | Use `tls: "full"` or valid cert |
| Handshake timeout | `flexible` | Origin expects TLS | Use `tls: "full"` |

**Debug:**
```bash
openssl s_client -connect app.example.com:443 -showcerts
```

### SMTP Reverse DNS

**Problem:** Email servers reject SMTP via Spectrum  
**Cause:** Spectrum IPs lack PTR (reverse DNS) records  
**Impact:** Many mail servers require valid rDNS for anti-spam

**Solution:**
- Outbound SMTP: NOT recommended through Spectrum
- Inbound SMTP: Use Cloudflare Email Routing
- Internal relay: Whitelist Spectrum IPs on destination

### Proxy Protocol Compatibility

**Problem:** Connection works but app behaves incorrectly  
**Cause:** Origin doesn't support Proxy Protocol

**Solution:**
1. Verify origin supports version (v1: widely supported, v2: HAProxy 1.5+/nginx 1.11+)
2. Test with `proxy_protocol: 'off'` first
3. Configure origin to parse headers

**nginx TCP:**
```nginx
stream {
    server {
        listen 22 proxy_protocol;
        proxy_pass backend:22;
    }
}
```

**HAProxy:**
```
frontend ft_ssh
    bind :22 accept-proxy
```

### Analytics Data Retention

**Problem:** Historical data not available  
**Cause:** Retention varies by plan

| Plan | Real-time | Historical |
|------|-----------|------------|
| Pro | Last hour | ❌ |
| Business | Last hour | Limited |
| Enterprise | Last hour | 90+ days |

**Solution:** Query within retention window or export to external system

### Enterprise-Only Features

**Problem:** Feature unavailable/errors  
**Cause:** Requires Enterprise plan

**Enterprise-only:**
- Port ranges (`tcp/25565-25575`)
- All TCP/UDP ports (Pro/Business: selected only)
- Extended analytics retention
- Advanced load balancing

### IPv6 Considerations

**Problem:** IPv6 clients can't connect or origin doesn't support IPv6  
**Solution:** Configure `edge_ips.connectivity`

```typescript
const app = await client.spectrum.apps.create({
  // ...
  edge_ips: {
    type: 'dynamic',
    connectivity: 'ipv4',  // Options: 'all', 'ipv4', 'ipv6'
  },
});
```

**Options:**
- `all`: Dual-stack (default, requires origin support both)
- `ipv4`: IPv4 only (use if origin lacks IPv6)
- `ipv6`: IPv6 only (rare)

## Limits

| Resource | Pro/Business | Enterprise |
|----------|--------------|------------|
| Max apps | ~10-15 | 100+ |
| Protocols | Selected | All TCP/UDP |
| Port ranges | ❌ | ✅ |
| Analytics | ~1 hour | 90+ days |

## See Also

- [patterns.md](patterns.md) - Protocol examples
- [configuration.md](configuration.md) - TLS/Proxy setup

## Common Use Cases

### 1. SSH Server Protection

**Terraform:**
```hcl
resource "cloudflare_spectrum_application" "ssh" {
  zone_id  = var.zone_id
  protocol = "tcp/22"

  dns {
    type = "CNAME"
    name = "ssh.example.com"
  }

  origin_direct      = ["tcp://10.0.1.5:22"]
  ip_firewall        = true
  argo_smart_routing = true
}
```

**Benefits:** Hide origin IP, DDoS protection, IP firewall, Argo reduces latency

### 2. Game Server

**TypeScript (Minecraft):**
```typescript
const app = await client.spectrum.apps.create({
  zone_id: 'your-zone-id',
  protocol: 'tcp/25565',
  dns: { type: 'CNAME', name: 'mc.example.com' },
  origin_direct: ['tcp://192.168.1.10:25565'],
  proxy_protocol: 'v1',  // Preserves player IPs
  argo_smart_routing: true,
});
```

**Benefits:** DDoS protection, hide origin IP, Proxy Protocol for player IPs/bans, Argo reduces latency

### 3. MQTT Broker

IoT device communication.

**TypeScript:**
```typescript
const mqttApp = await client.spectrum.apps.create({
  zone_id: 'your-zone-id',
  protocol: 'tcp/8883',  // Use 1883 for plain MQTT
  dns: { type: 'CNAME', name: 'mqtt.example.com' },
  origin_direct: ['tcp://mqtt-broker.internal:8883'],
  tls: 'full',  // Use 'off' for plain MQTT
});
```

**Benefits:** DDoS protection, hide broker IP, TLS termination at edge

### 4. SMTP Relay

Email submission (port 587). **WARNING**: See [gotchas.md](gotchas.md#smtp-reverse-dns)

**Terraform:**
```hcl
resource "cloudflare_spectrum_application" "smtp" {
  zone_id  = var.zone_id
  protocol = "tcp/587"

  dns {
    type = "CNAME"
    name = "smtp.example.com"
  }

  origin_direct = ["tcp://mail-server.internal:587"]
  tls           = "full"  # STARTTLS support
}
```

**Limitations:**
- Spectrum IPs lack reverse DNS (PTR records)
- Many mail servers reject without valid rDNS
- Best for internal/trusted relay only

### 5. Database Proxy

MySQL/PostgreSQL. **Use with caution** - security critical.

**PostgreSQL:**
```typescript
const postgresApp = await client.spectrum.apps.create({
  zone_id: 'your-zone-id',
  protocol: 'tcp/5432',
  dns: { type: 'CNAME', name: 'postgres.example.com' },
  origin_dns: { name: 'db-primary.internal.example.com' },
  origin_port: 5432,
  tls: 'strict',      // REQUIRED
  ip_firewall: true,  // REQUIRED
});
```

**MySQL:**
```hcl
resource "cloudflare_spectrum_application" "mysql" {
  zone_id  = var.zone_id
  protocol = "tcp/3306"

  dns {
    type = "CNAME"
    name = "mysql.example.com"
  }

  origin_dns {
    name = "mysql-primary.internal.example.com"
  }

  origin_port = 3306
  tls         = "strict"
  ip_firewall = true
}
```

**Security:**
- ALWAYS use `tls: "strict"`
- ALWAYS use `ip_firewall: true`
- Restrict to known IPs via zone firewall
- Use strong DB authentication
- Consider VPN or Cloudflare Access instead

### 6. RDP (Remote Desktop)

**Requires IP firewall.**

**Terraform:**
```hcl
resource "cloudflare_spectrum_application" "rdp" {
  zone_id  = var.zone_id
  protocol = "tcp/3389"

  dns {
    type = "CNAME"
    name = "rdp.example.com"
  }

  origin_direct = ["tcp://windows-server.internal:3389"]
  tls           = "off"       # RDP has own encryption
  ip_firewall   = true        # REQUIRED
}
```

**Security:** ALWAYS `ip_firewall: true`, whitelist admin IPs, RDP is DDoS/brute-force target

### 7. Multi-Origin Failover

High availability with load balancer.

**Terraform:**
```hcl
resource "cloudflare_load_balancer" "database_lb" {
  zone_id          = var.zone_id
  name             = "db-lb.example.com"
  default_pool_ids = [cloudflare_load_balancer_pool.db_primary.id]
  fallback_pool_id = cloudflare_load_balancer_pool.db_secondary.id
}

resource "cloudflare_load_balancer_pool" "db_primary" {
  name    = "db-primary-pool"
  origins { name = "db-1"; address = "192.0.2.1" }
  monitor = cloudflare_load_balancer_monitor.postgres_monitor.id
}

resource "cloudflare_load_balancer_pool" "db_secondary" {
  name    = "db-secondary-pool"
  origins { name = "db-2"; address = "192.0.2.2" }
  monitor = cloudflare_load_balancer_monitor.postgres_monitor.id
}

resource "cloudflare_load_balancer_monitor" "postgres_monitor" {
  type = "tcp"; port = 5432; interval = 30; timeout = 5
}

resource "cloudflare_spectrum_application" "postgres_ha" {
  zone_id     = var.zone_id
  protocol    = "tcp/5432"
  dns         { type = "CNAME"; name = "postgres.example.com" }
  origin_dns  { name = cloudflare_load_balancer.database_lb.name }
  origin_port = 5432
  tls         = "strict"
  ip_firewall = true
}
```

**Benefits:** Automatic failover, health monitoring, traffic distribution, zero-downtime deployments

## See Also

- [configuration.md](configuration.md) - Origin type setup
- [gotchas.md](gotchas.md) - Protocol limitations
- [api.md](api.md) - SDK reference

# Cloudflare Spectrum Skill Reference

## Overview

Cloudflare Spectrum provides security and acceleration for ANY TCP or UDP-based application. It's a global Layer 4 (L4) reverse proxy running on Cloudflare's edge nodes that routes MQTT, email, file transfer, version control, games, and more through Cloudflare to mask origins and protect from DDoS attacks.

**When to Use Spectrum**: When your protocol isn't HTTP/HTTPS (use Cloudflare proxy for HTTP). Spectrum handles everything else: SSH, gaming, databases, MQTT, SMTP, RDP, custom protocols.

## Plan Capabilities

| Capability | Pro/Business | Enterprise |
|------------|--------------|------------|
| TCP protocols | Selected ports only | All ports (1-65535) |
| UDP protocols | Selected ports only | All ports (1-65535) |
| Port ranges | ❌ | ✅ |
| Argo Smart Routing | ✅ | ✅ |
| IP Firewall | ✅ | ✅ |
| Load balancer origins | ✅ | ✅ |

## Decision Tree

**What are you trying to do?**

1. **Create/manage Spectrum app**
   - Via Dashboard → See [Cloudflare Dashboard](https://dash.cloudflare.com)
   - Via API → See [api.md](api.md) - REST endpoints
   - Via SDK → See [api.md](api.md) - TypeScript/Python/Go examples
   - Via IaC → See [configuration.md](configuration.md) - Terraform/Pulumi

2. **Protect specific protocol**
   - SSH → See [patterns.md](patterns.md#1-ssh-server-protection)
   - Gaming (Minecraft, etc) → See [patterns.md](patterns.md#2-game-server)
   - MQTT/IoT → See [patterns.md](patterns.md#3-mqtt-broker)
   - SMTP/Email → See [patterns.md](patterns.md#4-smtp-relay)
   - Database → See [patterns.md](patterns.md#5-database-proxy)
   - RDP → See [patterns.md](patterns.md#6-rdp-remote-desktop)

3. **Choose origin type**
   - Direct IP (single server) → See [configuration.md](configuration.md#direct-ip-origin)
   - CNAME (hostname) → See [configuration.md](configuration.md#cname-origin)
   - Load balancer (HA/failover) → See [configuration.md](configuration.md#load-balancer-origin)

## Reading Order

1. Start with [patterns.md](patterns.md) for your specific protocol
2. Then [configuration.md](configuration.md) for your origin type
3. Check [gotchas.md](gotchas.md) before going to production
4. Use [api.md](api.md) for programmatic access

## See Also

- [Cloudflare Docs](https://developers.cloudflare.com/spectrum/)

# Static Assets Binding API

Read the binding reference before implementing calls. Check the configured binding name and use the project's existing environment types.

| Task | Documentation |
|------|---------------|
| Forward a request or fetch a specific asset | [Runtime API reference](https://developers.cloudflare.com/workers/static-assets/binding/#runtime-api-reference) |
| Understand how binding requests apply HTML and fallback settings | [Binding fetch behavior](https://developers.cloudflare.com/workers/static-assets/binding/#runtime-api-reference) and [HTML handling](https://developers.cloudflare.com/workers/static-assets/routing/advanced/html-handling/) |
| Run authorization or transform content before serving | [Run your Worker script first](https://developers.cloudflare.com/workers/static-assets/routing/worker-script/#run-your-worker-script-first) |
| Inspect default MIME, cache, and validation headers | [Default headers](https://developers.cloudflare.com/workers/static-assets/headers/#default-headers) |
| Add or override response headers | [Custom headers and Worker-response caveat](https://developers.cloudflare.com/workers/static-assets/headers/#custom-headers) |
| Fetch assets imported by Vite | [Vite asset features](https://developers.cloudflare.com/workers/vite-plugin/reference/static-assets/#features) |

When selecting a different asset, construct a full URL using the incoming request as the base. Preserve the returned status and headers when transforming a response; do not collapse every unsuccessful response into an application 404.

See [configuration.md](configuration.md) for bindings and [patterns.md](patterns.md) for routing choices.

# Static Assets Configuration

Inspect the build script, output directory, Wrangler configuration, and any framework-generated deployment configuration first. Configure the files actually produced by the build, and identify which paths need Worker logic.

| Task | Documentation |
|------|---------------|
| Set the asset directory and exclude non-public files | [Directory](https://developers.cloudflare.com/workers/static-assets/binding/#directory) and [ignoring assets](https://developers.cloudflare.com/workers/static-assets/binding/#ignoring-assets) |
| Make assets available to Worker code | [Asset binding configuration](https://developers.cloudflare.com/workers/static-assets/binding/#binding) |
| Select paths that must run Worker logic before asset serving | [Worker-first configuration](https://developers.cloudflare.com/workers/static-assets/binding/#run_worker_first) |
| Configure SPA fallback and navigation behavior | [SPA configuration](https://developers.cloudflare.com/workers/static-assets/routing/single-page-application/#configuration) |
| Configure generated HTML and missing-page responses | [SSG configuration](https://developers.cloudflare.com/workers/static-assets/routing/static-site-generation/#configuration) |
| Choose canonical HTML URLs and trailing slash handling | [HTML handling modes](https://developers.cloudflare.com/workers/static-assets/routing/advanced/html-handling/) |
| Configure assets produced by the Cloudflare Vite plugin | [Vite asset configuration](https://developers.cloudflare.com/workers/vite-plugin/reference/static-assets/#configuration) |
| Configure static response headers or redirects | [Headers](https://developers.cloudflare.com/workers/static-assets/headers/) and [redirects](https://developers.cloudflare.com/workers/static-assets/redirects/) |

For Vite projects, inspect generated output before overriding asset paths. For protected routes, ensure that the selected routing configuration reaches the authorization logic before returning an asset. Use [patterns.md](patterns.md) to decide routing intent and [gotchas.md](gotchas.md) to verify it.

# Static Assets Gotchas

Compare the deployed build output and effective configuration with the exact request that failed. Test browser navigation separately from client-side fetches when investigating SPA routing.

| Symptom or decision | Documentation |
|---------------------|---------------|
| Asset missing or unexpected files uploaded | [Asset directory](https://developers.cloudflare.com/workers/static-assets/binding/#directory) and [ignore rules](https://developers.cloudflare.com/workers/static-assets/binding/#ignoring-assets) |
| Worker is bypassed for an asset or protected path | [Worker-first routing](https://developers.cloudflare.com/workers/static-assets/routing/worker-script/#run-your-worker-script-first) |
| Browser navigation returns HTML for an API path | [SPA navigation requests](https://developers.cloudflare.com/workers/static-assets/routing/single-page-application/#navigation-requests) and [advanced routing](https://developers.cloudflare.com/workers/static-assets/routing/single-page-application/#advanced-routing-control) |
| Unexpected redirects or trailing slashes | [HTML handling](https://developers.cloudflare.com/workers/static-assets/routing/advanced/html-handling/) |
| A missing route serves the wrong fallback | [SPA routing](https://developers.cloudflare.com/workers/static-assets/routing/single-page-application/) or [SSG custom 404 pages](https://developers.cloudflare.com/workers/static-assets/routing/static-site-generation/#custom-404-pages) |
| Cache behavior or custom headers differ from expectations | [Default and custom headers](https://developers.cloudflare.com/workers/static-assets/headers/) |
| Vite development and deployment behave differently | [Vite asset configuration and output](https://developers.cloudflare.com/workers/vite-plugin/reference/static-assets/) |
| Worker-first requests return 429 or affect cost | [Billing and limitations](https://developers.cloudflare.com/workers/static-assets/billing-and-limitations/) |
| Asset count or file size exceeds deployment limits | [Static asset platform limits](https://developers.cloudflare.com/workers/platform/limits/#static-assets) |
| Placement adds latency to asset requests | [Worker routing and placement caveat](https://developers.cloudflare.com/workers/static-assets/routing/worker-script/#run-your-worker-script-first) |

Verify a public asset, API endpoint, protected path, and missing URL against the routing intent. Inspect response status, redirect location, and cache headers before changing fallback or cache configuration. Return to [configuration.md](configuration.md) and [api.md](api.md) for the relevant settings and binding behavior.

# Static Assets Routing Patterns

Decide which requests need application logic before choosing configuration. Record expected behavior for public files, API paths, protected content, browser navigation, and missing URLs.

| Routing need | Documentation |
|--------------|---------------|
| Serve public assets with minimal Worker involvement | [Default Worker routing](https://developers.cloudflare.com/workers/static-assets/routing/worker-script/) |
| Combine a SPA with API routes | [Advanced SPA routing control](https://developers.cloudflare.com/workers/static-assets/routing/single-page-application/#advanced-routing-control) |
| Require authentication or transform assets before serving | [Run Worker before each request](https://developers.cloudflare.com/workers/static-assets/routing/worker-script/#run-worker-before-each-request) |
| Route an OAuth callback or selected dynamic paths to the Worker | [Run Worker first for selective paths](https://developers.cloudflare.com/workers/static-assets/routing/worker-script/#run-worker-first-for-selective-paths) |
| Select assets for locale or experiment variants | [Asset binding API](https://developers.cloudflare.com/workers/static-assets/binding/#runtime-api-reference) and [Worker-first routing](https://developers.cloudflare.com/workers/static-assets/binding/#run_worker_first) |
| Serve generated HTML with a custom 404 | [Custom 404 pages](https://developers.cloudflare.com/workers/static-assets/routing/static-site-generation/#custom-404-pages) |
| Apply static cache/security headers or redirects | [Custom headers](https://developers.cloudflare.com/workers/static-assets/headers/#custom-headers) and [redirects](https://developers.cloudflare.com/workers/static-assets/redirects/) |

Keep public asset paths eligible for direct serving when they do not need application logic. Ensure protected paths cannot bypass the authorization handler through an asset match or navigation fallback. For locale or experiment routing, map accepted variants to known build outputs and define a fallback explicitly.

Use [configuration.md](configuration.md) for setup and [gotchas.md](gotchas.md) for representative request checks.

# Workers Static Assets

Use Workers Static Assets for new static sites, SPAs, generated sites, and applications combining assets with server logic. Inspect the framework, build output, and existing deployment configuration before changing routing.

| Task | Documentation |
|------|---------------|
| Set up and deploy a static site or application | [Get started](https://developers.cloudflare.com/workers/static-assets/get-started/) |
| Choose configuration and an optional asset binding | [Configuration and bindings](https://developers.cloudflare.com/workers/static-assets/binding/) |
| Serve a client-rendered application | [SPA routing](https://developers.cloudflare.com/workers/static-assets/routing/single-page-application/) |
| Serve generated HTML and custom error pages | [SSG routing](https://developers.cloudflare.com/workers/static-assets/routing/static-site-generation/) |
| Use a full-stack framework | [Full-stack application guides](https://developers.cloudflare.com/workers/static-assets/routing/full-stack-application/) |
| Evaluate moving an existing Pages project | [Pages migration guide](https://developers.cloudflare.com/workers/static-assets/migration-guides/migrate-from-pages/) |

Do not choose a platform solely from the framework name. For an existing Pages project, inspect its current features and migration requirements before proposing a move.

## Reading Order

1. [configuration.md](configuration.md) — build output and routing configuration.
2. [api.md](api.md) — fetch assets and handle responses.
3. [patterns.md](patterns.md) — choose a routing design.
4. [gotchas.md](gotchas.md) — diagnose routing, caching, and deployment issues.

# Stream Live Streaming API

Choose the ingest protocol and viewer experience before creating live inputs. Read the current examples for returned endpoints and credentials instead of constructing publish URLs from an input ID.

| Task | Read |
|---|---|
| Create an RTMPS/SRT input and connect an encoder | [Start a live stream](https://developers.cloudflare.com/stream/stream-live/start-stream-live/) |
| Update inputs, recording, retention, or stream keys | [Manage live inputs](https://developers.cloudflare.com/stream/stream-live/start-stream-live/#manage-live-inputs) |
| Choose persistent-channel playback versus a particular video | [View by live input ID or video ID](https://developers.cloudflare.com/stream/stream-live/watch-live-stream/#view-by-live-input-id-or-video-id) |
| Find and replay recorded broadcasts | [Replay recordings](https://developers.cloudflare.com/stream/stream-live/replay-recordings/) |
| Forward broadcasts to external platforms | [Simulcasting configuration and limits](https://developers.cloudflare.com/stream/stream-live/simulcasting/) |
| Receive connection and disconnection notifications | [Live webhooks](https://developers.cloudflare.com/stream/stream-live/webhooks/) |
| Publish and play using WHIP/WHEP | [WebRTC requirements and endpoints](https://developers.cloudflare.com/stream/webrtc-beta/) and [browser integration](https://developers.cloudflare.com/stream/examples/browser-based-webrtc/) |
| Diagnose encoder, buffering, or latency problems | [Live troubleshooting](https://developers.cloudflare.com/stream/stream-live/troubleshooting/) |

Keep publishing credentials separate from viewer playback data. Decide whether the application stores a reusable live input, individual recording IDs, or both. Route recording processing events through [video webhooks](https://developers.cloudflare.com/stream/manage-video-library/using-webhooks/); live connection events have their own notification setup.

See [configuration.md](./configuration.md) for access decisions and [patterns.md](./patterns.md) for application state handling.

# Stream API Reference

Read the task-specific documentation for request schemas, SDK examples, and response fields. For live inputs and outputs, use [api-live.md](./api-live.md).

| Task | Read |
|---|---|
| Issue a one-time upload URL to an end user | [Direct creator uploads](https://developers.cloudflare.com/stream/uploading-videos/direct-creator-uploads/) |
| Resume uploads or handle unreliable connections | [Resumable uploads and requirements](https://developers.cloudflare.com/stream/uploading-videos/resumable-uploads/) |
| Import a video already hosted elsewhere | [Upload via link](https://developers.cloudflare.com/stream/uploading-videos/upload-via-link/) |
| Embed an iframe or React player | [Use the Stream Player](https://developers.cloudflare.com/stream/viewing-videos/using-the-stream-player/) |
| Use HLS/DASH with an existing player | [Use your own player](https://developers.cloudflare.com/stream/viewing-videos/using-own-player/) |
| Issue playback tokens | [Signed URLs and token-generation choices](https://developers.cloudflare.com/stream/viewing-videos/securing-your-stream/#three-ways-to-generate-signed-tokens) |
| Generate previews or downloadable files | [Thumbnails](https://developers.cloudflare.com/stream/viewing-videos/displaying-thumbnails/) and [downloads](https://developers.cloudflare.com/stream/viewing-videos/download-videos/) |
| Upload captions or generate them with Stream | [Add captions](https://developers.cloudflare.com/stream/edit-videos/adding-captions/) |
| Apply branding during upload | [Watermarks](https://developers.cloudflare.com/stream/edit-videos/applying-watermarks/) |
| Trim an on-demand video | [Video clipping](https://developers.cloudflare.com/stream/edit-videos/video-clipping/) |
| List and filter videos through the REST API | [List videos](https://developers.cloudflare.com/api/resources/stream/methods/list/) |
| Manage videos from a Worker | [Stream binding methods](https://developers.cloudflare.com/stream/manage-video-library/bindings/#methods) |
| React to encoding success or failure | [Video webhooks](https://developers.cloudflare.com/stream/manage-video-library/using-webhooks/) |

Keep the returned video UID associated with the application's owning user or record. Treat upload completion and playback readiness as separate application states; see [workflow decisions](./patterns.md).

# Stream Configuration

Inspect the existing runtime, API client or binding, authentication layer, player, and secret storage before adding Stream. Choose upload and viewing permissions from the application's requirements.

| Task | Read |
|---|---|
| Configure Stream in a Worker and use its binding | [Stream binding setup](https://developers.cloudflare.com/stream/manage-video-library/bindings/#setup) |
| Set creator upload constraints and metadata | [Direct creator uploads](https://developers.cloudflare.com/stream/uploading-videos/direct-creator-uploads/) |
| Choose a token endpoint, Worker binding, or signing key | [Token-generation choices](https://developers.cloudflare.com/stream/viewing-videos/securing-your-stream/#three-ways-to-generate-signed-tokens) |
| Require private playback or apply token restrictions | [Secure your Stream](https://developers.cloudflare.com/stream/viewing-videos/securing-your-stream/) |
| Restrict embedding origins | [Allowed origins](https://developers.cloudflare.com/stream/viewing-videos/securing-your-stream/#allowed-origins) |
| Configure processing notifications and their secret | [Video webhooks](https://developers.cloudflare.com/stream/manage-video-library/using-webhooks/) |
| Configure live recording or external destinations | [Live inputs](https://developers.cloudflare.com/stream/stream-live/start-stream-live/) and [simulcasting](https://developers.cloudflare.com/stream/stream-live/simulcasting/) |
| Choose player configuration or framework integration | [Stream Player](https://developers.cloudflare.com/stream/viewing-videos/using-the-stream-player/) |

Use the project's server-side secret handling for API tokens, signing keys, and webhook secrets. Identify the authorization check that permits issuing an upload URL or playback token; Stream configuration alone does not define the application's user entitlements.

See [api.md](./api.md), [api-live.md](./api-live.md), and [gotchas.md](./gotchas.md).

# Stream Gotchas

Identify whether the failure occurs during upload, encoding, authorization, playback, or live ingestion before changing configuration. Read the linked requirements and error guidance for that stage.

| Symptom or check | Read |
|---|---|
| Unsupported file or upload constraint failure | [Supported formats and upload requirements](https://developers.cloudflare.com/stream/uploading-videos/) |
| Large upload fails or restarts after interruption | [Resumable upload requirements](https://developers.cloudflare.com/stream/uploading-videos/resumable-uploads/#requirements) |
| Stream cannot fetch an imported video | [Upload via link requirements](https://developers.cloudflare.com/stream/uploading-videos/upload-via-link/) |
| Upload finished but video is not playable | [Upload progress tracking](https://developers.cloudflare.com/stream/uploading-videos/direct-creator-uploads/#track-upload-progress) and [processing error codes](https://developers.cloudflare.com/stream/manage-video-library/using-webhooks/#error-codes) |
| Private playback or embedding fails | [Signed tokens, restrictions, and allowed origins](https://developers.cloudflare.com/stream/viewing-videos/securing-your-stream/) |
| Webhook signature verification fails | [Verify the raw body and signature](https://developers.cloudflare.com/stream/manage-video-library/using-webhooks/#verify-webhook-authenticity) |
| A custom HLS/DASH player behaves incorrectly | [Own-player integration and manifest handling](https://developers.cloudflare.com/stream/viewing-videos/using-own-player/) |
| Live stream will not connect or playback buffers | [Live troubleshooting](https://developers.cloudflare.com/stream/stream-live/troubleshooting/) |
| Simulcast output does not behave as expected | [Output configuration and limits](https://developers.cloudflare.com/stream/stream-live/simulcasting/) |
| Estimate storage or delivery costs | [Stream pricing](https://developers.cloudflare.com/stream/pricing/) |

Check the actual response and current docs before adopting a retry policy or treating a video as ready. Keep credentials out of browser code and logs; use [configuration.md](./configuration.md) for access setup and [patterns.md](./patterns.md) for state decisions.

# Stream Patterns

Use the official workflow examples after identifying where the existing application handles authorization, video ownership, and processing state.

| Workflow | Read |
|---|---|
| Browser uploads through a server-issued URL | [Direct creator uploads](https://developers.cloudflare.com/stream/uploading-videos/direct-creator-uploads/) |
| Large or interruption-prone uploads | [Resumable uploads](https://developers.cloudflare.com/stream/uploading-videos/resumable-uploads/) |
| Update application state when processing completes | [Video webhooks](https://developers.cloudflare.com/stream/manage-video-library/using-webhooks/) and [upload progress tracking](https://developers.cloudflare.com/stream/uploading-videos/direct-creator-uploads/#track-upload-progress) |
| Verify an incoming processing notification | [Webhook authenticity](https://developers.cloudflare.com/stream/manage-video-library/using-webhooks/#verify-webhook-authenticity) |
| Embed playback in a React application | [Stream Player and framework integrations](https://developers.cloudflare.com/stream/viewing-videos/using-the-stream-player/) |
| Serve private videos | [Signed playback and signing examples](https://developers.cloudflare.com/stream/viewing-videos/securing-your-stream/) |
| Broadcast from a browser | [Browser-based WebRTC](https://developers.cloudflare.com/stream/examples/browser-based-webrtc/) |

## Application Decisions

- Check the user's permission before issuing upload URLs or playback tokens. Decide how uploaded video IDs map to application records and who may later view or delete them.
- Model uploading, processing, ready, and failed states in the UI. Use documented status and notification data; select retry and timeout policies to suit the application instead of assuming a fixed encoding deadline.
- Preserve the raw webhook request body for verification before applying state changes. Decide how the application's existing event handling reconciles notifications with stored video records.
- For live playback, decide whether viewers follow a channel across broadcasts or open a specific recording; see [live workflows](./api-live.md).

See [configuration.md](./configuration.md) for setup and [gotchas.md](./gotchas.md) for diagnosis.

# Cloudflare Stream

Use Stream for on-demand video upload and playback or live broadcasting. Start with the application's upload source, player, and access requirements, then read the corresponding documentation before writing code.

## Choose a Workflow

| Task | Read |
|---|---|
| Let users upload without exposing an API token | [Direct creator uploads](https://developers.cloudflare.com/stream/uploading-videos/direct-creator-uploads/) |
| Choose server upload, resumable upload, or import from a URL | [Upload methods and supported formats](https://developers.cloudflare.com/stream/uploading-videos/) |
| Embed playback or integrate an existing player | [Stream Player](https://developers.cloudflare.com/stream/viewing-videos/using-the-stream-player/) or [HLS/DASH players](https://developers.cloudflare.com/stream/viewing-videos/using-own-player/) |
| Restrict viewing to authorized users or embedding origins | [Secure your Stream](https://developers.cloudflare.com/stream/viewing-videos/securing-your-stream/) |
| Broadcast live, replay recordings, or simulcast | [Live workflow routing](./api-live.md) |
| Check usage, costs, or upload constraints | [Analytics](https://developers.cloudflare.com/stream/getting-analytics/), [pricing](https://developers.cloudflare.com/stream/pricing/), and [upload requirements](https://developers.cloudflare.com/stream/uploading-videos/) |

## In This Reference

- [configuration.md](./configuration.md): project setup and access decisions.
- [api.md](./api.md): upload, playback, editing, and library operations.
- [api-live.md](./api-live.md): live inputs, outputs, recording, and WebRTC.
- [patterns.md](./patterns.md): application workflow decisions.
- [gotchas.md](./gotchas.md): troubleshooting routes.

# Tail Workers APIs

Fetch the handler reference for current event shapes and language examples instead of maintaining local interface definitions.

| Task | Documentation |
| --- | --- |
| Implement the handler and understand asynchronous processing | [Tail handler syntax](https://developers.cloudflare.com/workers/runtime-apis/handlers/tail/#syntax) |
| Interpret execution outcomes, logs, exceptions, and timestamps | [Tail handler event reference](https://developers.cloudflare.com/workers/runtime-apis/handlers/tail/) |
| Inspect request fields and redaction behavior | [TailRequest](https://developers.cloudflare.com/workers/runtime-apis/handlers/tail/#tailrequest) |
| Interpret events from dynamic dispatch and user Workers | [Handler parameters](https://developers.cloudflare.com/workers/runtime-apis/handlers/tail/#parameters) |
| Write aggregated metrics | [Analytics Engine from Tail Workers](https://developers.cloudflare.com/workers/observability/logs/tail-workers/#use-analytics-engine-for-aggregated-metrics) |

Choose the event fields required by the destination and preserve the distinction between execution outcome and HTTP response status. Review what data may be retained before bypassing redaction; the documented heuristics are not a complete application privacy policy. See [patterns.md](./patterns.md) for filter design.

# Tail Workers Configuration

Identify the producer Worker and consumer Worker as separate deployment resources before changing their configuration.

| Task | Documentation |
| --- | --- |
| Connect a producer to a Tail Worker | [Configure Tail Workers](https://developers.cloudflare.com/workers/observability/logs/tail-workers/#configure-tail-workers) |
| Configure environments and resource bindings | [Wrangler configuration](https://developers.cloudflare.com/workers/wrangler/configuration/) |
| Configure destination credentials | [Secrets](https://developers.cloudflare.com/workers/configuration/secrets/) |
| Use built-in telemetry destinations instead | [Exporting OpenTelemetry Data](https://developers.cloudflare.com/workers/observability/exporting-opentelemetry-data/) |
| Check availability and billing model | [Tail Workers](https://developers.cloudflare.com/workers/observability/logs/tail-workers/) |
| Check runtime capacity | [Workers limits](https://developers.cloudflare.com/workers/platform/limits/) |
| Check the handler required by the consumer | [Tail handler](https://developers.cloudflare.com/workers/runtime-apis/handlers/tail/) |

Confirm which project owns each deployment and which environment the destination belongs to. Verify a known producer request reaches the intended destination before expanding coverage. See [api.md](./api.md) for payload handling and [gotchas.md](./gotchas.md) for investigation.

# Tail Workers Troubleshooting

Trace one known producer invocation through consumer execution and destination receipt to locate the failure.

| Task | Documentation |
| --- | --- |
| Consumer receives no events or producer setup fails | [Configure Tail Workers](https://developers.cloudflare.com/workers/observability/logs/tail-workers/#configure-tail-workers) |
| Async processing or event interpretation is incorrect | [Tail handler](https://developers.cloudflare.com/workers/runtime-apis/handlers/tail/) |
| Filters confuse HTTP responses with execution failures | [TailItems](https://developers.cloudflare.com/workers/runtime-apis/handlers/tail/#tailitems) |
| Request data is missing or unexpectedly retained | [TailRequest](https://developers.cloudflare.com/workers/runtime-apis/handlers/tail/#tailrequest) |
| Inspect producer and consumer execution | [Real-time logs](https://developers.cloudflare.com/workers/observability/logs/real-time-logs/) |
| Reconsider a custom exporter at higher volume | [Exporting OpenTelemetry Data](https://developers.cloudflare.com/workers/observability/exporting-opentelemetry-data/) |

Verify producer configuration, consumer identity, and destination credentials independently. Include both handled error responses and thrown exceptions when checking filters. Make destination failures observable without exposing the payload or credentials in diagnostic logs. Do not infer retention, retries, or batch guarantees from an example.

See [api.md](./api.md) for data contracts and [patterns.md](./patterns.md) for destination decisions.

# Tail Workers Processing Decisions

Define the destination contract and required event coverage before writing transformations.

| Task | Documentation |
| --- | --- |
| Export supported telemetry without a custom consumer | [Exporting OpenTelemetry Data](https://developers.cloudflare.com/workers/observability/exporting-opentelemetry-data/) |
| Process and forward custom execution events | [Tail Workers](https://developers.cloudflare.com/workers/observability/logs/tail-workers/) |
| Aggregate metrics rather than retain individual events | [Analytics Engine from Tail Workers](https://developers.cloudflare.com/workers/observability/logs/tail-workers/#use-analytics-engine-for-aggregated-metrics) |
| Choose filters based on execution and request fields | [Tail handler](https://developers.cloudflare.com/workers/runtime-apis/handlers/tail/) |
| Review sensitive request data handling | [TailRequest](https://developers.cloudflare.com/workers/runtime-apis/handlers/tail/#tailrequest) |

Decide which producers, routes, failures, and successful requests each destination needs. If sampling is appropriate, specify how it affects the questions the data must answer. Minimize retained fields, test transformation and serialization against representative events, and check destination rejection behavior.

Keep producer identity when combining events from multiple Workers. Define how delivery failures become visible before adding fallback storage or batching infrastructure. See [configuration.md](./configuration.md) for deployment ownership.

# Cloudflare Tail Workers

Use Tail Workers when execution events need custom processing. Fetch the current documentation before implementing handlers, configuration, or integrations.

| Task | Documentation |
| --- | --- |
| Decide whether custom processing is needed | [Tail Workers](https://developers.cloudflare.com/workers/observability/logs/tail-workers/) |
| Export logs and traces to an observability destination | [Exporting OpenTelemetry Data](https://developers.cloudflare.com/workers/observability/exporting-opentelemetry-data/) |
| Inspect a deployment interactively | [Real-time logs](https://developers.cloudflare.com/workers/observability/logs/real-time-logs/) |
| Implement the consumer | [Tail handler](https://developers.cloudflare.com/workers/runtime-apis/handlers/tail/) |

Before adding a Tail Worker, check whether built-in OpenTelemetry export meets the destination’s needs. Use the Tail Workers guide for the tradeoff, then identify the custom filtering or transformation that remains necessary.

## In This Reference

- [configuration.md](./configuration.md) — producer, consumer, destination, and environment setup
- [api.md](./api.md) — event fields, execution outcomes, and redaction
- [patterns.md](./patterns.md) — destination and filtering decisions
- [gotchas.md](./gotchas.md) — connection, data, and delivery investigation

See [observability](../observability/README.md) for broader logging and tracing choices.

# Terraform Data Sources Reference

Query existing Cloudflare resources to reference in your configurations.

## v5 Data Source Names

| v4 Name | v5 Name | Notes |
|---------|---------|-------|
| `cloudflare_record` | `cloudflare_dns_record` | |
| `cloudflare_worker_script` | `cloudflare_workers_script` | Note: plural |
| `cloudflare_access_*` | `cloudflare_zero_trust_*` | Access → Zero Trust |

## Zone Data Sources

```hcl
# Get zone by name
data "cloudflare_zone" "example" {
  name = "example.com"
}

# Use in resources
resource "cloudflare_dns_record" "www" {
  zone_id = data.cloudflare_zone.example.id
  name = "www"
  # ...
}
```

## Account Data Sources

```hcl
# List all accounts
data "cloudflare_accounts" "main" {
  name = "My Account"
}

# Use account ID
resource "cloudflare_worker_script" "api" {
  account_id = data.cloudflare_accounts.main.accounts[0].id
  # ...
}
```

## Worker Data Sources

```hcl
# Get existing worker script (v5: cloudflare_workers_script)
data "cloudflare_workers_script" "existing" {
  account_id = var.account_id
  name = "existing-worker"
}

# Reference in service bindings
resource "cloudflare_workers_script" "consumer" {
  service_binding {
    name = "UPSTREAM"
    service = data.cloudflare_workers_script.existing.name
  }
}
```

## KV Data Sources

```hcl
# Get KV namespace
data "cloudflare_workers_kv_namespace" "existing" {
  account_id = var.account_id
  namespace_id = "abc123"
}

# Use in worker binding
resource "cloudflare_workers_script" "api" {
  kv_namespace_binding {
    name = "KV"
    namespace_id = data.cloudflare_workers_kv_namespace.existing.id
  }
}
```

## Lists Data Source

```hcl
# Get IP lists for WAF rules
data "cloudflare_list" "blocked_ips" {
  account_id = var.account_id
  name = "blocked_ips"
}
```

## IP Ranges Data Source

```hcl
# Get Cloudflare IP ranges (for firewall rules)
data "cloudflare_ip_ranges" "cloudflare" {}

output "ipv4_cidrs" {
  value = data.cloudflare_ip_ranges.cloudflare.ipv4_cidr_blocks
}

output "ipv6_cidrs" {
  value = data.cloudflare_ip_ranges.cloudflare.ipv6_cidr_blocks
}

# Use in security group rules (AWS example)
resource "aws_security_group_rule" "allow_cloudflare" {
  type = "ingress"
  from_port = 443
  to_port = 443
  protocol = "tcp"
  cidr_blocks = data.cloudflare_ip_ranges.cloudflare.ipv4_cidr_blocks
  security_group_id = aws_security_group.web.id
}
```

## Common Patterns

### Import ID Formats

| Resource | Import ID Format |
|----------|------------------|
| `cloudflare_zone` | `<zone-id>` |
| `cloudflare_dns_record` | `<zone-id>/<record-id>` |
| `cloudflare_workers_script` | `<account-id>/<script-name>` |
| `cloudflare_workers_kv_namespace` | `<account-id>/<namespace-id>` |
| `cloudflare_r2_bucket` | `<account-id>/<bucket-name>` |
| `cloudflare_d1_database` | `<account-id>/<database-id>` |
| `cloudflare_pages_project` | `<account-id>/<project-name>` |

```bash
# Example: Import DNS record
terraform import cloudflare_dns_record.example <zone-id>/<record-id>
```

### Reference Across Modules

```hcl
# modules/worker/main.tf
data "cloudflare_zone" "main" {
  name = var.domain
}

resource "cloudflare_worker_route" "api" {
  zone_id = data.cloudflare_zone.main.id
  pattern = "api.${var.domain}/*"
  script_name = cloudflare_worker_script.api.name
}
```

### Output Important Values

```hcl
output "zone_id" {
  value = cloudflare_zone.main.id
  description = "Zone ID for DNS management"
}

output "worker_url" {
  value = "https://${cloudflare_worker_domain.api.hostname}"
  description = "Worker API endpoint"
}

output "kv_namespace_id" {
  value = cloudflare_workers_kv_namespace.app.id
  sensitive = false
}

output "name_servers" {
  value = cloudflare_zone.main.name_servers
  description = "Name servers for domain registration"
}
```

## See Also

- [README](./README.md) - Provider setup
- [Configuration Reference](./configuration.md) - All resource types
- [Patterns](./patterns.md) - Architecture patterns
- [Troubleshooting](./gotchas.md) - Common issues

# Terraform Configuration Reference

Complete resource configurations for Cloudflare infrastructure.

## Zone & DNS

```hcl
# Zone + settings
resource "cloudflare_zone" "example" { account = { id = var.account_id }; name = "example.com"; type = "full" }
resource "cloudflare_zone_settings_override" "example" {
  zone_id = cloudflare_zone.example.id
  settings { ssl = "strict"; always_use_https = "on"; min_tls_version = "1.2"; tls_1_3 = "on"; http3 = "on" }
}

# DNS records (A, CNAME, MX, TXT)
resource "cloudflare_dns_record" "www" {
  zone_id = cloudflare_zone.example.id; name = "www"; content = "192.0.2.1"; type = "A"; proxied = true
}
resource "cloudflare_dns_record" "mx" {
  for_each = { "10" = "mail1.example.com", "20" = "mail2.example.com" }
  zone_id = cloudflare_zone.example.id; name = "@"; content = each.value; type = "MX"; priority = each.key
}
```

## Workers

### Simple Pattern (Legacy - Still Works)

```hcl
resource "cloudflare_workers_script" "api" {
  account_id = var.account_id; name = "api-worker"; content = file("worker.js")
  module = true; compatibility_date = "2025-01-01"
  kv_namespace_binding { name = "KV"; namespace_id = cloudflare_workers_kv_namespace.cache.id }
  r2_bucket_binding { name = "BUCKET"; bucket_name = cloudflare_r2_bucket.assets.name }
  d1_database_binding { name = "DB"; database_id = cloudflare_d1_database.app.id }
  secret_text_binding { name = "SECRET"; text = var.secret }
}
```

### Gradual Rollouts (Recommended for Production)

```hcl
resource "cloudflare_worker" "api" { account_id = var.account_id; name = "api-worker" }
resource "cloudflare_worker_version" "api_v1" {
  account_id = var.account_id; worker_name = cloudflare_worker.api.name
  content = file("worker.js"); content_sha256 = filesha256("worker.js")
  compatibility_date = "2025-01-01"
  bindings {
    kv_namespace { name = "KV"; namespace_id = cloudflare_workers_kv_namespace.cache.id }
    r2_bucket { name = "BUCKET"; bucket_name = cloudflare_r2_bucket.assets.name }
  }
}
resource "cloudflare_workers_deployment" "api" {
  account_id = var.account_id; worker_name = cloudflare_worker.api.name
  versions { version_id = cloudflare_worker_version.api_v1.id; percentage = 100 }
}
```

### Worker Binding Types (v5)

| Binding | Attribute | Example |
|---------|-----------|---------|
| KV | `kv_namespace_binding` | `{ name = "KV", namespace_id = "..." }` |
| R2 | `r2_bucket_binding` | `{ name = "BUCKET", bucket_name = "..." }` |
| D1 | `d1_database_binding` | `{ name = "DB", database_id = "..." }` |
| Service | `service_binding` | `{ name = "AUTH", service = "auth-worker" }` |
| Secret | `secret_text_binding` | `{ name = "API_KEY", text = "..." }` |
| Queue | `queue_binding` | `{ name = "QUEUE", queue_name = "..." }` |
| Vectorize | `vectorize_binding` | `{ name = "INDEX", index_name = "..." }` |
| Hyperdrive | `hyperdrive_binding` | `{ name = "DB", id = "..." }` |
| AI | `ai_binding` | `{ name = "AI" }` |
| Browser | `browser_binding` | `{ name = "BROWSER" }` |
| Analytics | `analytics_engine_binding` | `{ name = "ANALYTICS", dataset = "..." }` |
| mTLS | `mtls_certificate_binding` | `{ name = "CERT", certificate_id = "..." }` |

### Routes & Triggers

```hcl
resource "cloudflare_worker_route" "api" {
  zone_id = cloudflare_zone.example.id; pattern = "api.example.com/*"
  script_name = cloudflare_workers_script.api.name
}
resource "cloudflare_worker_cron_trigger" "task" {
  account_id = var.account_id; script_name = cloudflare_workers_script.api.name
  schedules = ["*/5 * * * *"]
}
```

## Storage (KV, R2, D1)

```hcl
# KV
resource "cloudflare_workers_kv_namespace" "cache" { account_id = var.account_id; title = "cache" }
resource "cloudflare_workers_kv" "config" {
  account_id = var.account_id; namespace_id = cloudflare_workers_kv_namespace.cache.id
  key_name = "config"; value = jsonencode({ version = "1.0" })
}

# R2
resource "cloudflare_r2_bucket" "assets" { account_id = var.account_id; name = "assets"; location = "WNAM" }

# D1 (migrations via wrangler) & Queues
resource "cloudflare_d1_database" "app" { account_id = var.account_id; name = "app-db" }
resource "cloudflare_queue" "events" { account_id = var.account_id; name = "events-queue" }
```

## Pages

```hcl
resource "cloudflare_pages_project" "site" {
  account_id = var.account_id; name = "site"; production_branch = "main"
  deployment_configs {
    production {
      compatibility_date = "2025-01-01"
      environment_variables = { NODE_ENV = "production" }
      kv_namespaces = { KV = cloudflare_workers_kv_namespace.cache.id }
      d1_databases = { DB = cloudflare_d1_database.app.id }
    }
  }
  build_config { build_command = "npm run build"; destination_dir = "dist" }
  source { type = "github"; config { owner = "org"; repo_name = "site"; production_branch = "main" }}
}

resource "cloudflare_pages_domain" "custom" {
  account_id = var.account_id; project_name = cloudflare_pages_project.site.name; domain = "site.example.com"
}
```

## Rulesets (WAF, Redirects, Cache)

```hcl
# WAF
resource "cloudflare_ruleset" "waf" {
  zone_id = cloudflare_zone.example.id; name = "WAF"; kind = "zone"; phase = "http_request_firewall_custom"
  rules { action = "block"; enabled = true; expression = "(cf.client.bot) and not (cf.verified_bot)" }
}

# Redirects
resource "cloudflare_ruleset" "redirects" {
  zone_id = cloudflare_zone.example.id; name = "Redirects"; kind = "zone"; phase = "http_request_dynamic_redirect"
  rules {
    action = "redirect"; enabled = true; expression = "(http.request.uri.path eq \"/old\")"
    action_parameters { from_value { status_code = 301; target_url { value = "https://example.com/new" }}}
  }
}

# Cache rules
resource "cloudflare_ruleset" "cache" {
  zone_id = cloudflare_zone.example.id; name = "Cache"; kind = "zone"; phase = "http_request_cache_settings"
  rules {
    action = "set_cache_settings"; enabled = true; expression = "(http.request.uri.path matches \"\\.(jpg|png|css|js)$\")"
    action_parameters { cache = true; edge_ttl { mode = "override_origin"; default = 86400 }}
  }
}
```

## Load Balancers

```hcl
resource "cloudflare_load_balancer_monitor" "http" {
  account_id = var.account_id; type = "http"; path = "/health"; interval = 60; timeout = 5
}
resource "cloudflare_load_balancer_pool" "api" {
  account_id = var.account_id; name = "api-pool"; monitor = cloudflare_load_balancer_monitor.http.id
  origins { name = "api-1"; address = "192.0.2.1" }
  origins { name = "api-2"; address = "192.0.2.2" }
}
resource "cloudflare_load_balancer" "api" {
  zone_id = cloudflare_zone.example.id; name = "api.example.com"
  default_pool_ids = [cloudflare_load_balancer_pool.api.id]; steering_policy = "geo"
}
```

## Access (Zero Trust)

```hcl
resource "cloudflare_access_application" "admin" {
  account_id = var.account_id; name = "Admin"; domain = "admin.example.com"; type = "self_hosted"
  session_duration = "24h"; allowed_idps = [cloudflare_access_identity_provider.github.id]
}
resource "cloudflare_access_policy" "allow" {
  account_id = var.account_id; application_id = cloudflare_access_application.admin.id
  name = "Allow"; decision = "allow"; precedence = 1
  include { email = ["admin@example.com"] }
}
resource "cloudflare_access_identity_provider" "github" {
  account_id = var.account_id; name = "GitHub"; type = "github"
  config { client_id = var.github_id; client_secret = var.github_secret }
}
```

## See Also

- [README](./README.md) - Provider setup
- [API](./api.md) - Data sources
- [Patterns](./patterns.md) - Use cases
- [Troubleshooting](./gotchas.md) - Issues

# Terraform Troubleshooting & Best Practices

Common issues, security considerations, and best practices.

## State Drift Issues

Some resources have known state drift. Add lifecycle blocks to prevent perpetual diffs:

| Resource | Drift Attributes | Workaround |
|----------|------------------|------------|
| `cloudflare_pages_project` | `deployment_configs.*` | `ignore_changes = [deployment_configs]` |
| `cloudflare_workers_script` | secrets returned as REDACTED | `ignore_changes = [secret_text_binding]` |
| `cloudflare_load_balancer` | `adaptive_routing`, `random_steering` | `ignore_changes = [adaptive_routing, random_steering]` |
| `cloudflare_workers_kv` | special chars in keys (< 5.16.0) | Upgrade to 5.16.0+ |

```hcl
# Example: Ignore secret drift
resource "cloudflare_workers_script" "api" {
  account_id = var.account_id
  name = "api-worker"
  content = file("worker.js")
  secret_text_binding { name = "API_KEY"; text = var.api_key }
  
  lifecycle {
    ignore_changes = [secret_text_binding]
  }
}
```

## v5 Breaking Changes

Provider v5 is current (auto-generated from OpenAPI). v4→v5 has breaking changes:

**Resource Renames:**

| v4 Resource | v5 Resource | Notes |
|-------------|-------------|-------|
| `cloudflare_record` | `cloudflare_dns_record` | |
| `cloudflare_worker_script` | `cloudflare_workers_script` | Note: plural |
| `cloudflare_worker_*` | `cloudflare_workers_*` | All worker resources |
| `cloudflare_access_*` | `cloudflare_zero_trust_*` | Access → Zero Trust |

**Attribute Changes:**

| v4 Attribute | v5 Attribute | Resources |
|--------------|--------------|-----------|
| `zone` | `name` | zone |
| `account_id` | `account.id` | zone (object syntax) |
| `key` | `key_name` | KV |
| `location_hint` | `location` | R2 |

**State Migration:**

```bash
# Rename resources in state after v5 upgrade
terraform state mv cloudflare_record.example cloudflare_dns_record.example
terraform state mv cloudflare_worker_script.api cloudflare_workers_script.api
```

## Resource-Specific Gotchas

### R2 Location Case Sensitivity

**Problem:** Terraform creates R2 bucket but fails on subsequent applies  
**Cause:** Location must be UPPERCASE  
**Solution:** Use `WNAM`, `ENAM`, `WEUR`, `EEUR`, `APAC` (not `wnam`, `enam`, etc.)

```hcl
resource "cloudflare_r2_bucket" "assets" {
  account_id = var.account_id
  name = "assets"
  location = "WNAM"  # UPPERCASE required
}
```

### KV Special Characters (< 5.16.0)

**Problem:** Keys with `+`, `#`, `%` cause encoding issues  
**Cause:** URL encoding bug in provider < 5.16.0  
**Solution:** Upgrade to 5.16.0+ or avoid special chars in keys

### D1 Migrations

**Problem:** Terraform creates database but schema is empty  
**Cause:** Terraform only creates D1 resource, not schema  
**Solution:** Run migrations via wrangler after Terraform apply

```bash
# After terraform apply
wrangler d1 migrations apply <db-name>
```

### Worker Script Size Limit

**Problem:** Worker deployment fails with "script too large"  
**Cause:** Worker script + dependencies exceed 10 MB limit  
**Solution:** Use code splitting, external dependencies, or minification

### Pages Project Drift

**Problem:** Pages project shows perpetual diff on `deployment_configs`  
**Cause:** Cloudflare API adds default values not in Terraform state  
**Solution:** Add lifecycle ignore block (see State Drift table above)

## Common Errors

### "Error: couldn't find resource"

**Cause:** Resource was deleted outside Terraform  
**Solution:** Import resource back into state with `terraform import cloudflare_zone.example <zone-id>` or remove from state with `terraform state rm cloudflare_zone.example`

### "409 Conflict on worker deployment"

**Cause:** Worker being deployed by both Terraform and wrangler simultaneously  
**Solution:** Choose one deployment method; if using Terraform, remove wrangler deployments

### "DNS record already exists"

**Cause:** Existing DNS record not imported into Terraform state  
**Solution:** Find record ID in Cloudflare dashboard and import with `terraform import cloudflare_dns_record.example <zone-id>/<record-id>`

### "Invalid provider configuration"

**Cause:** API token missing, invalid, or lacking required permissions  
**Solution:** Set `CLOUDFLARE_API_TOKEN` environment variable or check token permissions in dashboard

### "State locking errors"

**Cause:** Multiple concurrent Terraform runs or stale lock from crashed process  
**Solution:** Remove stale lock with `terraform force-unlock <lock-id>` (use with caution)

## Limits

| Resource | Limit | Notes |
|----------|-------|-------|
| API token rate limit | Varies by plan | Use `api_client_logging = true` to debug
| Worker script size | 10 MB | Includes all dependencies
| KV keys per namespace | Unlimited | Pay per operation
| R2 storage | Unlimited | Pay per GB
| D1 databases | 50,000 per account | Free tier: 10
| Pages projects | 500 per account | 100 for free accounts
| DNS records | 3,500 per zone | Free plan

## See Also

- [README](./README.md) - Provider setup
- [Configuration](./configuration.md) - Resources
- [API](./api.md) - Data sources
- [Patterns](./patterns.md) - Use cases
- Provider docs: https://registry.terraform.io/providers/cloudflare/cloudflare/latest/docs

# Terraform Patterns & Use Cases

Architecture patterns, multi-environment setups, and real-world use cases.

## Recommended Directory Structure

```
terraform/
├── environments/
│   ├── production/
│   │   ├── main.tf
│   │   └── terraform.tfvars
│   └── staging/
│       ├── main.tf
│       └── terraform.tfvars
├── modules/
│   ├── zone/
│   ├── worker/
│   └── dns/
└── shared/          # Shared resources across envs
    └── main.tf
```

**Note:** Cloudflare recommends avoiding modules for provider resources due to v5 auto-generation complexity. Prefer environment directories + shared state instead.

## Multi-Environment Setup

```hcl
# Directory: environments/{production,staging}/main.tf + modules/{zone,worker,pages}
module "zone" {
  source = "../../modules/zone"; account_id = var.account_id; zone_name = "example.com"; environment = "production"
}
module "api_worker" {
  source = "../../modules/worker"; account_id = var.account_id; zone_id = module.zone.zone_id
  name = "api-worker-prod"; script = file("../../workers/api.js"); environment = "production"
}
```

## R2 State Backend

```hcl
terraform {
  backend "s3" {
    bucket = "terraform-state"
    key = "cloudflare.tfstate"
    region = "auto"
    endpoints = { s3 = "https://<account_id>.r2.cloudflarestorage.com" }
    skip_credentials_validation = true
    skip_region_validation = true
    skip_requesting_account_id = true
    skip_metadata_api_check = true
    skip_s3_checksum = true
  }
}
```

## Worker with All Bindings

```hcl
locals { worker_name = "full-stack-worker" }
resource "cloudflare_workers_kv_namespace" "app" { account_id = var.account_id; title = "${local.worker_name}-kv" }
resource "cloudflare_r2_bucket" "app" { account_id = var.account_id; name = "${local.worker_name}-bucket" }
resource "cloudflare_d1_database" "app" { account_id = var.account_id; name = "${local.worker_name}-db" }

resource "cloudflare_worker_script" "app" {
  account_id = var.account_id; name = local.worker_name; content = file("worker.js"); module = true
  compatibility_date = "2025-01-01"
  kv_namespace_binding { name = "KV"; namespace_id = cloudflare_workers_kv_namespace.app.id }
  r2_bucket_binding { name = "BUCKET"; bucket_name = cloudflare_r2_bucket.app.name }
  d1_database_binding { name = "DB"; database_id = cloudflare_d1_database.app.id }
  secret_text_binding { name = "API_KEY"; text = var.api_key }
}
```

## Wrangler Integration

**CRITICAL**: Wrangler and Terraform must NOT manage same resources.

**Terraform**: Zones, DNS, security rules, Access, load balancers, worker deployments (CI/CD), KV/R2/D1 resource creation  
**Wrangler**: Local dev (`wrangler dev`), manual deploys, D1 migrations, KV bulk ops, log streaming (`wrangler tail`)

### CI/CD Pattern

```hcl
# Terraform creates infrastructure
resource "cloudflare_workers_kv_namespace" "app" { account_id = var.account_id; title = "app-kv" }
resource "cloudflare_d1_database" "app" { account_id = var.account_id; name = "app-db" }
output "kv_namespace_id" { value = cloudflare_workers_kv_namespace.app.id }
output "d1_database_id" { value = cloudflare_d1_database.app.id }
```

```yaml
# GitHub Actions: terraform apply → envsubst wrangler.jsonc.template → wrangler deploy
- run: terraform apply -auto-approve
- run: |
    export KV_NAMESPACE_ID=$(terraform output -raw kv_namespace_id)
    envsubst < wrangler.jsonc.template > wrangler.jsonc
- run: wrangler deploy
```

## Use Cases

### Static Site + API Worker

```hcl
resource "cloudflare_pages_project" "frontend" {
  account_id = var.account_id; name = "frontend"; production_branch = "main"
  build_config { build_command = "npm run build"; destination_dir = "dist" }
}
resource "cloudflare_worker_script" "api" {
  account_id = var.account_id; name = "api"; content = file("api-worker.js")
  d1_database_binding { name = "DB"; database_id = cloudflare_d1_database.api_db.id }
}
resource "cloudflare_dns_record" "frontend" {
  zone_id = cloudflare_zone.main.id; name = "app"; content = cloudflare_pages_project.frontend.subdomain; type = "CNAME"; proxied = true
}
resource "cloudflare_worker_route" "api" {
  zone_id = cloudflare_zone.main.id; pattern = "api.example.com/*"; script_name = cloudflare_worker_script.api.name
}
```

### Multi-Region Load Balancing

```hcl
resource "cloudflare_load_balancer_pool" "us" {
  account_id = var.account_id; name = "us-pool"; monitor = cloudflare_load_balancer_monitor.http.id
  origins { name = "us-east"; address = var.us_east_ip }
}
resource "cloudflare_load_balancer_pool" "eu" {
  account_id = var.account_id; name = "eu-pool"; monitor = cloudflare_load_balancer_monitor.http.id
  origins { name = "eu-west"; address = var.eu_west_ip }
}
resource "cloudflare_load_balancer" "global" {
  zone_id = cloudflare_zone.main.id; name = "api.example.com"; steering_policy = "geo"
  default_pool_ids = [cloudflare_load_balancer_pool.us.id]
  region_pools { region = "WNAM"; pool_ids = [cloudflare_load_balancer_pool.us.id] }
  region_pools { region = "WEU"; pool_ids = [cloudflare_load_balancer_pool.eu.id] }
}
```

### Secure Admin with Access

```hcl
resource "cloudflare_pages_project" "admin" { account_id = var.account_id; name = "admin"; production_branch = "main" }
resource "cloudflare_access_application" "admin" {
  account_id = var.account_id; name = "Admin"; domain = "admin.example.com"; type = "self_hosted"; session_duration = "24h"
  allowed_idps = [cloudflare_access_identity_provider.google.id]
}
resource "cloudflare_access_policy" "allow" {
  account_id = var.account_id; application_id = cloudflare_access_application.admin.id
  name = "Allow admins"; decision = "allow"; precedence = 1; include { email = var.admin_emails }
}
```

### Reusable Module

```hcl
# modules/cloudflare-zone/main.tf
variable "account_id" { type = string }; variable "domain" { type = string }; variable "ssl_mode" { default = "strict" }
resource "cloudflare_zone" "main" { account = { id = var.account_id }; name = var.domain }
resource "cloudflare_zone_settings_override" "main" {
  zone_id = cloudflare_zone.main.id; settings { ssl = var.ssl_mode; always_use_https = "on" }
}
output "zone_id" { value = cloudflare_zone.main.id }

# Usage: module "prod" { source = "./modules/cloudflare-zone"; account_id = var.account_id; domain = "example.com" }
```

## See Also

- [README](./README.md) - Provider setup
- [Configuration Reference](./configuration.md) - All resource types
- [API Reference](./api.md) - Data sources
- [Troubleshooting](./gotchas.md) - Best practices, common issues

# Cloudflare Terraform Provider

**Expert guidance for Cloudflare Terraform Provider - infrastructure as code for Cloudflare resources.**

## Core Principles

- **Provider-first**: Use Terraform provider for ALL infrastructure - never mix with wrangler.jsonc for the same resources
- **State management**: Always use remote state (S3, Terraform Cloud, etc.) for team environments
- **Modular architecture**: Create reusable modules for common patterns (zones, workers, pages)
- **Version pinning**: Always pin provider version with `~>` for predictable upgrades
- **Secret management**: Use variables + environment vars for sensitive data - never hardcode API tokens

## Provider Version

| Version | Status | Notes |
|---------|--------|-------|
| 5.x | Current | Auto-generated from OpenAPI, breaking changes from v4 |
| 4.x | Legacy | Manual maintenance, deprecated |

**Critical:** v5 renamed many resources (`cloudflare_record` → `cloudflare_dns_record`, `cloudflare_worker_*` → `cloudflare_workers_*`). See [gotchas.md](./gotchas.md#v5-breaking-changes) for migration details.

## Provider Setup

### Basic Configuration

```hcl
terraform {
  required_version = ">= 1.0"
  
  required_providers {
    cloudflare = {
      source  = "cloudflare/cloudflare"
      version = "~> 5.15.0"
    }
  }
}

provider "cloudflare" {
  api_token = var.cloudflare_api_token  # or CLOUDFLARE_API_TOKEN env var
}
```

### Authentication Methods (priority order)

1. **API Token** (RECOMMENDED): `api_token` or `CLOUDFLARE_API_TOKEN`
   - Create: Dashboard → My Profile → API Tokens
   - Scope to specific accounts/zones for security
   
2. **Global API Key** (LEGACY): `api_key` + `api_email` or `CLOUDFLARE_API_KEY` + `CLOUDFLARE_EMAIL`
   - Less secure, use tokens instead
   
3. **User Service Key**: `user_service_key` for Origin CA certificates



## Quick Reference: Common Commands

```bash
terraform init          # Initialize provider
terraform plan          # Plan changes
terraform apply         # Apply changes
terraform destroy       # Destroy resources
terraform import cloudflare_zone.example <zone-id>  # Import existing
terraform state list    # List resources in state
terraform output        # Show outputs
terraform fmt -recursive  # Format code
terraform validate      # Validate configuration
```

## Import Existing Resources

Use cf-terraforming to generate configs from existing Cloudflare resources:

```bash
# Install
brew install cloudflare/cloudflare/cf-terraforming

# Generate HCL from existing resources
cf-terraforming generate --resource-type cloudflare_dns_record --zone <zone-id>

# Import into Terraform state
cf-terraforming import --resource-type cloudflare_dns_record --zone <zone-id>
```

## Reading Order

1. Start with [README.md](./README.md) for provider setup and authentication
2. Review [configuration.md](./configuration.md) for resource configurations
3. Check [api.md](./api.md) for data sources and existing resource queries
4. See [patterns.md](./patterns.md) for multi-environment and CI/CD patterns
5. Read [gotchas.md](./gotchas.md) for state drift, v5 breaking changes, and troubleshooting

## In This Reference
- [configuration.md](./configuration.md) - Resources for zones, DNS, workers, KV, R2, D1, Pages, rulesets
- [api.md](./api.md) - Data sources for existing resources
- [patterns.md](./patterns.md) - Architecture patterns, multi-env setup, CI/CD integration
- [gotchas.md](./gotchas.md) - Common issues, security, best practices

## See Also
- [pulumi](../pulumi/) - Alternative IaC tool for Cloudflare
- [wrangler](https://developers.cloudflare.com/workers/wrangler/) - CLI deployment alternative
- [workers](https://developers.cloudflare.com/workers/) - Worker runtime documentation

# Tunnel APIs and Commands

Fetch current operation schemas, permissions, and examples before automating tunnel changes.

| Task | Documentation |
| --- | --- |
| Create a tunnel, configure ingress, and create application DNS through the API | [Setup](https://developers.cloudflare.com/tunnel/setup/) |
| List existing tunnels and inspect response fields | [List Cloudflare Tunnels API](https://developers.cloudflare.com/api/resources/zero_trust/subresources/tunnels/subresources/cloudflared/methods/list/) |
| Retrieve and rotate tunnel tokens | [Tunnel tokens](https://developers.cloudflare.com/tunnel/advanced/tunnel-tokens/) |
| Manage local tunnels using the CLI | [Useful commands](https://developers.cloudflare.com/tunnel/advanced/local-management/tunnel-useful-commands/) |
| Configure public DNS and routing behavior | [Routing](https://developers.cloudflare.com/tunnel/routing/) |
| Select private-network integration and its setup guide | [Integrations](https://developers.cloudflare.com/tunnel/integrations/) |

Identify the account, tunnel ID, and management mode before a write. Review existing routes before replacing configuration; distinguish tunnel lifecycle operations from DNS changes. Use [patterns.md](./patterns.md) when infrastructure as code owns these resources.

# Tunnel Configuration

Read the documentation for the existing management mode before changing routes or credentials.

| Task | Documentation |
| --- | --- |
| Configure remotely-managed tunnels | [Setup](https://developers.cloudflare.com/tunnel/setup/) |
| Edit local ingress rules, service mappings, and validate matching | [Configuration file](https://developers.cloudflare.com/tunnel/advanced/local-management/configuration-file/) |
| Configure origin TLS, HTTP, and connection behavior | [Origin parameters](https://developers.cloudflare.com/tunnel/advanced/origin-parameters/) |
| Configure runtime flags and service arguments | [Run parameters](https://developers.cloudflare.com/tunnel/advanced/run-parameters/) |
| Manage remote tunnel tokens and rotation | [Tunnel tokens](https://developers.cloudflare.com/tunnel/advanced/tunnel-tokens/) |
| Choose service protocols and DNS routing | [Routing](https://developers.cloudflare.com/tunnel/routing/) |

Confirm which configuration source the running process uses, then review the routes affected by the change. Match origin settings to the actual service and certificate rather than copying settings from a different deployment. See [networking.md](./networking.md) for connectivity and [patterns.md](./patterns.md) for rollout decisions.

# Tunnel Troubleshooting

Capture the tunnel status, failing route, management mode, and cloudflared logs before changing settings.

| Task | Documentation |
| --- | --- |
| Tunnel fails to connect or reports an error | [Troubleshooting](https://developers.cloudflare.com/tunnel/troubleshooting/) |
| Tunnel is healthy but an HTTPS application fails or redirects | [HTTPS origins](https://developers.cloudflare.com/tunnel/troubleshooting/https-origins/) |
| Inspect connection health and application diagnostics | [Monitoring](https://developers.cloudflare.com/tunnel/monitoring/) |
| Check local configuration and rule matching | [Configuration file](https://developers.cloudflare.com/tunnel/advanced/local-management/configuration-file/) |
| Connections behave unexpectedly after token rotation | [Tunnel tokens](https://developers.cloudflare.com/tunnel/advanced/tunnel-tokens/) |
| Check replica capacity or firewall requirements | [Configuration](https://developers.cloudflare.com/tunnel/configuration/) |
| Update an existing installation | [Update cloudflared](https://developers.cloudflare.com/tunnel/downloads/update-cloudflared/) |

Separate tunnel health from origin availability. Check the service address, protocol, and certificate before relaxing verification. Scope operational changes to the intended tunnel and replicas; follow the documented rotation and upgrade sequence instead of stopping every cloudflared process on a host. See [networking.md](./networking.md) for connectivity checks.

# Tunnel Networking

Investigate the connection from cloudflared to Cloudflare separately from the connection to the origin and the client access path.

| Task | Documentation |
| --- | --- |
| Determine required egress ports and destinations | [Firewall rules](https://developers.cloudflare.com/tunnel/configuration/#firewall-rules) |
| Diagnose DNS, QUIC, or TCP connectivity | [Connection errors](https://developers.cloudflare.com/tunnel/troubleshooting/#connection-errors) |
| Configure transport and runtime options | [Run parameters](https://developers.cloudflare.com/tunnel/advanced/run-parameters/) |
| Check service protocols and client requirements | [Routing](https://developers.cloudflare.com/tunnel/routing/) |
| Set up private-network access with Cloudflare One | [Connect private networks with cloudflared](https://developers.cloudflare.com/cloudflare-one/networks/connectors/cloudflare-tunnel/private-net/cloudflared/) |
| Inspect tunnel health, logs, and metrics | [Monitoring](https://developers.cloudflare.com/tunnel/monitoring/) |

Test from the machine or container running cloudflared. Compare the actual firewall policy with the current documented destinations; do not infer transport ports from the origin protocol. For private access, follow the linked Cloudflare One setup for routes and device-client configuration instead of reusing public-hostname instructions. See [gotchas.md](./gotchas.md) for origin failures.

# Tunnel Deployment Decisions

Choose deployment ownership and availability requirements before adapting an example.

| Task | Documentation |
| --- | --- |
| Install a remotely-managed tunnel on a host or in Docker | [Setup](https://developers.cloudflare.com/tunnel/setup/) |
| Deploy cloudflared inside a cluster | [Kubernetes](https://developers.cloudflare.com/tunnel/deployment-guides/kubernetes/) |
| Manage tunnel infrastructure declaratively | [Terraform](https://developers.cloudflare.com/tunnel/deployment-guides/terraform/) |
| Deploy replicas and check current capacity | [Replicas and high availability](https://developers.cloudflare.com/tunnel/configuration/#replicas-and-high-availability) |
| Choose redundancy or explicit traffic steering | [Routing](https://developers.cloudflare.com/tunnel/routing/) |
| Plan upgrades for the existing installation method | [Update cloudflared](https://developers.cloudflare.com/tunnel/downloads/update-cloudflared/) |
| Add authentication or private connectivity | [Integrations](https://developers.cloudflare.com/tunnel/integrations/) |

Establish which service each replica can reach and how configuration and credentials reach each host. Decide whether simple redundancy meets the requirement or whether health-based routing needs a load balancer. Verify replacement replicas before retiring existing ones, and include application-level checks in the rollout. See [api.md](./api.md) for programmatic operations.

# Cloudflare Tunnel

Use Tunnel to connect origin services to Cloudflare. Inspect the existing tunnel, management mode, and intended audience before choosing a setup. Fetch current docs for commands, configuration, and limits.

| Task | Documentation |
| --- | --- |
| Create a remotely-managed tunnel or a temporary development tunnel | [Setup](https://developers.cloudflare.com/tunnel/setup/) |
| Maintain a tunnel managed through local files | [Create a locally-managed tunnel](https://developers.cloudflare.com/tunnel/advanced/local-management/create-local-tunnel/) |
| Publish an application and check protocol requirements | [Routing](https://developers.cloudflare.com/tunnel/routing/) |
| Choose private networking, Workers VPC, or Access integration | [Integrations](https://developers.cloudflare.com/tunnel/integrations/) |

Decide whether the goal is a public application, authenticated private access, or connectivity from a Worker. Then identify who owns configuration and how it will be deployed; multiple environments alone do not require local management.

## In This Reference

- [configuration.md](./configuration.md) — management mode, ingress, and origin settings
- [networking.md](./networking.md) — firewall, connectivity, and private-network investigation
- [api.md](./api.md) — programmatic setup and tunnel operations
- [patterns.md](./patterns.md) — deployment and availability decisions
- [gotchas.md](./gotchas.md) — troubleshooting and operational checks

# TURN API Reference

Complete API documentation for Cloudflare TURN service credentials and key management.

## Authentication

All endpoints require Cloudflare API token with "Calls Write" permission.

Base URL: `https://api.cloudflare.com/client/v4`

## TURN Key Management

### List TURN Keys

```
GET /accounts/{account_id}/calls/turn_keys
```

### Get TURN Key Details

```
GET /accounts/{account_id}/calls/turn_keys/{key_id}
```

### Create TURN Key

```
POST /accounts/{account_id}/calls/turn_keys
Content-Type: application/json

{
  "name": "my-turn-key"
}
```

**Response includes**:
- `uid`: Key identifier
- `key`: The actual secret key (only returned on creation—save immediately)
- `name`: Human-readable name
- `created`: ISO 8601 timestamp
- `modified`: ISO 8601 timestamp

### Update TURN Key

```
PUT /accounts/{account_id}/calls/turn_keys/{key_id}
Content-Type: application/json

{
  "name": "updated-name"
}
```

### Delete TURN Key

```
DELETE /accounts/{account_id}/calls/turn_keys/{key_id}
```

## Generate Temporary Credentials

```
POST https://rtc.live.cloudflare.com/v1/turn/keys/{key_id}/credentials/generate
Authorization: Bearer {key_secret}
Content-Type: application/json

{
  "ttl": 86400
}
```

### Credential Constraints

| Parameter | Min | Max | Default | Notes |
|-----------|-----|-----|---------|-------|
| ttl | 1 | 172800 (48hrs) | varies | API rejects values >172800 |

**CRITICAL**: Maximum TTL is 48 hours (172800 seconds). API will reject requests exceeding this limit.

### Response Schema

```json
{
  "iceServers": {
    "urls": [
      "stun:stun.cloudflare.com:3478",
      "turn:turn.cloudflare.com:3478?transport=udp",
      "turn:turn.cloudflare.com:3478?transport=tcp",
      "turn:turn.cloudflare.com:53?transport=udp",
      "turn:turn.cloudflare.com:80?transport=tcp",
      "turns:turn.cloudflare.com:5349?transport=tcp",
      "turns:turn.cloudflare.com:443?transport=tcp"
    ],
    "username": "1738035200:user123",
    "credential": "base64encodedhmac=="
  }
}
```

**Port 53 Warning**: Filter port 53 URLs for browser clients—blocked by Chrome/Firefox. See [gotchas.md](./gotchas.md#using-port-53-in-browsers).

## Revoke Credentials

```
POST https://rtc.live.cloudflare.com/v1/turn/keys/{key_id}/credentials/revoke
Authorization: Bearer {key_secret}
Content-Type: application/json

{
  "username": "1738035200:user123"
}
```

**Response**: 204 No Content

Billing stops immediately. Active connection drops after short delay (~seconds).

## TypeScript Types

```typescript
interface CloudflareTURNConfig {
  keyId: string;
  keySecret: string;
  ttl?: number; // Max 172800 (48 hours)
}

interface TURNCredentialsRequest {
  ttl?: number; // Max 172800 seconds
}

interface TURNCredentialsResponse {
  iceServers: {
    urls: string[];
    username: string;
    credential: string;
  };
}

interface RTCIceServer {
  urls: string | string[];
  username?: string;
  credential?: string;
  credentialType?: "password";
}

interface TURNKeyResponse {
  uid: string;
  key: string; // Only present on creation
  name: string;
  created: string;
  modified: string;
}
```

## Validation Function

```typescript
function validateRTCIceServer(obj: unknown): obj is RTCIceServer {
  if (!obj || typeof obj !== 'object') {
    return false;
  }

  const server = obj as Record<string, unknown>;

  if (typeof server.urls !== 'string' && !Array.isArray(server.urls)) {
    return false;
  }

  if (server.username && typeof server.username !== 'string') {
    return false;
  }

  if (server.credential && typeof server.credential !== 'string') {
    return false;
  }

  return true;
}
```

## Type-Safe Credential Generation

```typescript
async function fetchTURNServers(
  config: CloudflareTURNConfig
): Promise<RTCIceServer[]> {
  // Validate TTL constraint
  const ttl = config.ttl ?? 3600;
  if (ttl > 172800) {
    throw new Error('TTL cannot exceed 172800 seconds (48 hours)');
  }

  const response = await fetch(
    `https://rtc.live.cloudflare.com/v1/turn/keys/${config.keyId}/credentials/generate`,
    {
      method: 'POST',
      headers: {
        'Authorization': `Bearer ${config.keySecret}`,
        'Content-Type': 'application/json'
      },
      body: JSON.stringify({ ttl })
    }
  );

  if (!response.ok) {
    throw new Error(`TURN credential generation failed: ${response.status}`);
  }

  const data = await response.json();
  
  // Filter port 53 for browser clients
  const filteredUrls = data.iceServers.urls.filter(
    (url: string) => !url.includes(':53')
  );

  const iceServers = [
    { urls: 'stun:stun.cloudflare.com:3478' },
    {
      urls: filteredUrls,
      username: data.iceServers.username,
      credential: data.iceServers.credential,
      credentialType: 'password' as const
    }
  ];

  // Validate before returning
  if (!iceServers.every(validateRTCIceServer)) {
    throw new Error('Invalid ICE server configuration received');
  }

  return iceServers;
}
```

## See Also

- [configuration.md](./configuration.md) - Worker setup, environment variables
- [patterns.md](./patterns.md) - Implementation examples using these APIs
- [gotchas.md](./gotchas.md) - Security best practices, common mistakes

# TURN Configuration

Setup and configuration for Cloudflare TURN service in Workers and applications.

## Environment Variables

```bash
# .env
CLOUDFLARE_ACCOUNT_ID=your_account_id
CLOUDFLARE_API_TOKEN=your_api_token
TURN_KEY_ID=your_turn_key_id
TURN_KEY_SECRET=your_turn_key_secret
```

Validate with zod:

```typescript
import { z } from 'zod';

const envSchema = z.object({
  CLOUDFLARE_ACCOUNT_ID: z.string().min(1),
  CLOUDFLARE_API_TOKEN: z.string().min(1),
  TURN_KEY_ID: z.string().min(1),
  TURN_KEY_SECRET: z.string().min(1)
});

export const config = envSchema.parse(process.env);
```

## wrangler.jsonc

```jsonc
{
  "name": "turn-credentials-api",
  "main": "src/index.ts",
  "compatibility_date": "2025-01-01",
  "vars": {
    "TURN_KEY_ID": "your-turn-key-id"  // Non-sensitive, can be in vars
  },
  "env": {
    "production": {
      "kv_namespaces": [
        {
          "binding": "CREDENTIALS_CACHE",
          "id": "your-kv-namespace-id"
        }
      ]
    }
  }
}
```

**Store secrets separately**:
```bash
wrangler secret put TURN_KEY_SECRET
```

## Cloudflare Worker Integration

### Worker Binding Types

```typescript
interface Env {
  TURN_KEY_ID: string;
  TURN_KEY_SECRET: string;
  CREDENTIALS_CACHE?: KVNamespace;
}

export default {
  async fetch(request: Request, env: Env): Promise<Response> {
    // See patterns.md for implementation
  }
}
```

### Basic Worker Example

```typescript
export default {
  async fetch(request: Request, env: Env): Promise<Response> {
    if (request.url.endsWith('/turn-credentials')) {
      // Validate client auth
      const authHeader = request.headers.get('Authorization');
      if (!authHeader) {
        return new Response('Unauthorized', { status: 401 });
      }

      const response = await fetch(
        `https://rtc.live.cloudflare.com/v1/turn/keys/${env.TURN_KEY_ID}/credentials/generate`,
        {
          method: 'POST',
          headers: {
            'Authorization': `Bearer ${env.TURN_KEY_SECRET}`,
            'Content-Type': 'application/json'
          },
          body: JSON.stringify({ ttl: 3600 })
        }
      );

      if (!response.ok) {
        return new Response('Failed to generate credentials', { status: 500 });
      }

      const data = await response.json();

      // Filter port 53 for browser clients
      const filteredUrls = data.iceServers.urls.filter(
        (url: string) => !url.includes(':53')
      );

      return Response.json({
        iceServers: [
          { urls: 'stun:stun.cloudflare.com:3478' },
          {
            urls: filteredUrls,
            username: data.iceServers.username,
            credential: data.iceServers.credential
          }
        ]
      });
    }

    return new Response('Not found', { status: 404 });
  }
};
```

## IP Allowlisting (Enterprise/Firewall)

For strict firewalls, allowlist these IPs for `turn.cloudflare.com`:

| Type | Address | Protocol |
|------|---------|----------|
| IPv4 | 141.101.90.1/32 | All |
| IPv4 | 162.159.207.1/32 | All |
| IPv6 | 2a06:98c1:3200::1/128 | All |
| IPv6 | 2606:4700:48::1/128 | All |

**IMPORTANT**: These IPs may change with 14-day notice. Monitor DNS:

```bash
# Check A and AAAA records
dig turn.cloudflare.com A
dig turn.cloudflare.com AAAA
```

Set up automated monitoring to detect IP changes and update allowlists within 14 days.

## IPv6 Support

- **Client-to-TURN**: Both IPv4 and IPv6 supported
- **Relay addresses**: IPv4 only (no RFC 6156 support)
- **TCP relaying**: Not supported (RFC 6062)

Clients can connect via IPv6, but relayed traffic uses IPv4 addresses.

## TLS Configuration

### Supported TLS Versions
- TLS 1.1
- TLS 1.2
- TLS 1.3

### Recommended Ciphers (TLS 1.3)
- AEAD-AES128-GCM-SHA256
- AEAD-AES256-GCM-SHA384
- AEAD-CHACHA20-POLY1305-SHA256

### Recommended Ciphers (TLS 1.2)
- ECDHE-ECDSA-AES128-GCM-SHA256
- ECDHE-RSA-AES128-GCM-SHA256
- ECDHE-RSA-AES128-SHA (also TLS 1.1)
- AES128-GCM-SHA256

## See Also

- [api.md](./api.md) - TURN key creation, credential generation API
- [patterns.md](./patterns.md) - Full Worker implementation patterns
- [gotchas.md](./gotchas.md) - Security best practices, troubleshooting

# TURN Gotchas & Troubleshooting

Common mistakes, security best practices, and troubleshooting for Cloudflare TURN.

## Quick Reference

| Issue | Solution | Details |
|-------|----------|---------|
| Credentials not working | Check TTL ≤ 48hrs | [See Troubleshooting](#issue-turn-credentials-not-working) |
| Connection drops after ~48hrs | Implement credential refresh | [See Connection Drops](#issue-connection-drops-after-48-hours) |
| Port 53 fails in browser | Filter server-side | [See Port 53](#using-port-53-in-browsers) |
| High packet loss | Check rate limits | [See Rate Limits](#limits-per-turn-allocation) |
| Connection fails after maintenance | Implement ICE restart | [See ICE Restart](#ice-restart-required-scenarios) |

## Critical Constraints

| Constraint | Value | Consequence if Violated |
|------------|-------|-------------------------|
| Max credential TTL | 48 hours (172800s) | API rejects request |
| Credential revocation delay | ~seconds | Billing stops immediately, connection drops shortly |
| IP allowlist update window | 14 days (if IPs change) | Connection fails if IPs change |
| Packet rate | 5-10k pps per allocation | Packet drops |
| Data rate | 50-100 Mbps per allocation | Packet drops |
| Unique IP rate | >5 new IPs/sec | Packet drops |

## Limits Per TURN Allocation

**Per user** (not account-wide):

- **IP addresses**: >5 new unique IPs per second
- **Packet rate**: 5-10k packets per second (inbound/outbound)
- **Data rate**: 50-100 Mbps (inbound/outbound)
- **MTU**: No specific limit
- **Burst rates**: Higher than documented

Exceeding limits results in **packet drops**.

## Common Mistakes

### Setting TTL > 48 hours

```typescript
// ❌ BAD: API will reject
const creds = await generate({ ttl: 604800 });  // 7 days

// ✅ GOOD:
const creds = await generate({ ttl: 86400 });   // 24 hours
```

### Hardcoding IPs without monitoring

```typescript
// ❌ BAD: IPs can change with 14-day notice
const iceServers = [{ urls: 'turn:141.101.90.1:3478' }];

// ✅ GOOD: Use DNS
const iceServers = [{ urls: 'turn:turn.cloudflare.com:3478' }];
```

### Using port 53 in browsers

```typescript
// ❌ BAD: Blocked by Chrome/Firefox
urls: ['turn:turn.cloudflare.com:53']

// ✅ GOOD: Filter port 53
urls: urls.filter(url => !url.includes(':53'))
```

### Not handling credential expiry

```typescript
// ❌ BAD: Credentials expire but call continues → connection drops
const creds = await fetchCreds();
const pc = new RTCPeerConnection({ iceServers: creds });

// ✅ GOOD: Refresh before expiry
setInterval(() => refreshCredentials(pc), 3000000);  // 50 min
```

### Missing ICE restart support

```typescript
// ❌ BAD: No recovery from TURN maintenance
pc.addEventListener('iceconnectionstatechange', () => {
  console.log('State changed:', pc.iceConnectionState);
});

// ✅ GOOD: Implement ICE restart
pc.addEventListener('iceconnectionstatechange', async () => {
  if (pc.iceConnectionState === 'failed') {
    await refreshCredentials(pc);
    pc.restartIce();
  }
});
```

### Exposing TURN key secret client-side

```typescript
// ❌ BAD: Secret exposed to client
const secret = 'your-turn-key-secret';
const response = await fetch(`https://rtc.live.cloudflare.com/v1/turn/...`, {
  headers: { 'Authorization': `Bearer ${secret}` }
});

// ✅ GOOD: Generate credentials server-side
const response = await fetch('/api/turn-credentials');
```

## ICE Restart Required Scenarios

These events require ICE restart (see [patterns.md](./patterns.md#ice-restart-pattern)):

1. **TURN server maintenance** (occasional on Cloudflare's network)
2. **Network topology changes** (anycast routing changes)
3. **Credential refresh** during long sessions (>1 hour)
4. **Connection failure** (iceConnectionState === 'failed')

Implement in all production apps:

```typescript
pc.addEventListener('iceconnectionstatechange', async () => {
  if (pc.iceConnectionState === 'failed' || 
      pc.iceConnectionState === 'disconnected') {
    await refreshTURNCredentials(pc);
    pc.restartIce();
    const offer = await pc.createOffer({ iceRestart: true });
    await pc.setLocalDescription(offer);
    // Send offer to peer via signaling...
  }
});
```

Reference: [RFC 8445 Section 2.4](https://datatracker.ietf.org/doc/html/rfc8445#section-2.4)

## Security Checklist

- [ ] Credentials generated server-side only (never client-side)
- [ ] TURN_KEY_SECRET in wrangler secrets, not vars
- [ ] TTL ≤ expected session duration (and ≤ 48 hours)
- [ ] Rate limiting on credential generation endpoint
- [ ] Client authentication before issuing credentials
- [ ] Credential revocation API for compromised sessions
- [ ] No hardcoded IPs (or DNS monitoring in place)
- [ ] Port 53 filtered for browser clients

## Troubleshooting

### Issue: TURN credentials not working

**Check:**
- Key ID and secret are correct
- Credentials haven't expired (check TTL)
- TTL doesn't exceed 172800 seconds (48 hours)
- Server can reach rtc.live.cloudflare.com
- Network allows outbound HTTPS

**Solution:**
```typescript
// Validate before using
if (ttl > 172800) {
  throw new Error('TTL cannot exceed 48 hours');
}
```

### Issue: Slow connection establishment

**Solutions:**
- Ensure proper ICE candidate gathering
- Check network latency to Cloudflare edge
- Verify firewall allows WebRTC ports (3478, 5349, 443)
- Consider using TURN over TLS (port 443) for corporate networks

### Issue: High packet loss

**Check:**
- Not exceeding rate limits (5-10k pps)
- Not exceeding bandwidth limits (50-100 Mbps)
- Not connecting to too many unique IPs (>5/sec)
- Client network quality

### Issue: Connection drops after ~48 hours

**Cause**: Credentials expired (48hr max)

**Solution**: 
- Set TTL to expected session duration
- Implement credential refresh with setConfiguration()
- Use ICE restart if connection fails

```typescript
// Refresh credentials before expiry
const refreshInterval = ttl * 1000 - 60000; // 1 min early
setInterval(async () => {
  await refreshTURNCredentials(pc);
}, refreshInterval);
```

### Issue: Port 53 URLs in browser fail silently

**Cause**: Chrome/Firefox block port 53

**Solution**: Filter port 53 URLs server-side:

```typescript
const filtered = urls.filter(url => !url.includes(':53'));
```

### Issue: Hardcoded IPs stop working

**Cause**: Cloudflare changed IP addresses (14-day notice)

**Solution**: 
- Use DNS hostnames (`turn.cloudflare.com`)
- Monitor DNS changes with automated alerts
- Update allowlists within 14 days if using IP allowlisting

## Cost Optimization

1. Use appropriate TTLs (don't over-provision)
2. Implement credential caching
3. Set `iceTransportPolicy: 'all'` to try direct first (use `'relay'` only when necessary)
4. Monitor bandwidth usage
5. Free when used with Cloudflare Calls SFU

## See Also

- [api.md](./api.md) - Credential generation API, revocation
- [configuration.md](./configuration.md) - IP allowlisting, monitoring
- [patterns.md](./patterns.md) - ICE restart, credential refresh patterns

# TURN Implementation Patterns

Production-ready patterns for implementing Cloudflare TURN in WebRTC applications.

## Prerequisites

Before implementing these patterns, ensure you have:
- TURN key created: see [api.md#create-turn-key](./api.md#create-turn-key)
- Worker configured: see [configuration.md#cloudflare-worker-integration](./configuration.md#cloudflare-worker-integration)

## Basic TURN Configuration (Browser)

```typescript
interface RTCIceServer {
  urls: string | string[];
  username?: string;
  credential?: string;
  credentialType?: "password" | "oauth";
}

async function getTURNConfig(): Promise<RTCIceServer[]> {
  const response = await fetch('/api/turn-credentials');
  const data = await response.json();
  
  return [
    {
      urls: 'stun:stun.cloudflare.com:3478'
    },
    {
      urls: [
        'turn:turn.cloudflare.com:3478?transport=udp',
        'turn:turn.cloudflare.com:3478?transport=tcp',
        'turns:turn.cloudflare.com:5349?transport=tcp',
        'turns:turn.cloudflare.com:443?transport=tcp'
      ],
      username: data.username,
      credential: data.credential,
      credentialType: 'password'
    }
  ];
}

// Use in RTCPeerConnection
const iceServers = await getTURNConfig();
const peerConnection = new RTCPeerConnection({ iceServers });
```

## Port Selection Strategy

Recommended order for browser clients:

1. **3478/udp** (primary, lowest latency)
2. **3478/tcp** (fallback for UDP-blocked networks)
3. **5349/tls** (corporate firewalls, most reliable)
4. **443/tls** (alternate TLS port, firewall-friendly)

**Avoid port 53**—blocked by Chrome and Firefox.

```typescript
function filterICEServersForBrowser(urls: string[]): string[] {
  return urls
    .filter(url => !url.includes(':53'))  // Remove port 53
    .sort((a, b) => {
      // Prioritize UDP over TCP over TLS
      if (a.includes('transport=udp')) return -1;
      if (b.includes('transport=udp')) return 1;
      if (a.includes('transport=tcp') && !a.startsWith('turns:')) return -1;
      if (b.includes('transport=tcp') && !b.startsWith('turns:')) return 1;
      return 0;
    });
}
```

## Credential Refresh (Mid-Session)

When credentials expire during long calls:

```typescript
async function refreshTURNCredentials(pc: RTCPeerConnection): Promise<void> {
  const newCreds = await fetch('/turn-credentials').then(r => r.json());
  const config = pc.getConfiguration();
  config.iceServers = newCreds.iceServers;
  pc.setConfiguration(config);
  // Note: setConfiguration() does NOT trigger ICE restart
  // Combine with restartIce() if connection fails
}

// Auto-refresh before expiry
setInterval(async () => {
  await refreshTURNCredentials(peerConnection);
}, 3000000);  // 50 minutes if TTL is 1 hour
```

## ICE Restart Pattern

After network change, TURN server maintenance, or credential expiry:

```typescript
pc.addEventListener('iceconnectionstatechange', async () => {
  if (pc.iceConnectionState === 'failed') {
    console.warn('ICE connection failed, restarting...');
    
    // Refresh credentials
    await refreshTURNCredentials(pc);
    
    // Trigger ICE restart
    pc.restartIce();
    const offer = await pc.createOffer({ iceRestart: true });
    await pc.setLocalDescription(offer);
    
    // Send offer to peer via signaling channel...
  }
});
```

## Credentials Caching Pattern

```typescript
class TURNCredentialsManager {
  private creds: { username: string; credential: string; urls: string[]; expiresAt: number; } | null = null;

  async getCredentials(keyId: string, keySecret: string): Promise<RTCIceServer[]> {
    const now = Date.now();
    
    if (this.creds && this.creds.expiresAt > now) {
      return this.buildIceServers(this.creds);
    }

    const ttl = 3600;
    if (ttl > 172800) throw new Error('TTL max 48hrs');

    const res = await fetch(
      `https://rtc.live.cloudflare.com/v1/turn/keys/${keyId}/credentials/generate`,
      {
        method: 'POST',
        headers: { 'Authorization': `Bearer ${keySecret}`, 'Content-Type': 'application/json' },
        body: JSON.stringify({ ttl })
      }
    );

    const data = await res.json();
    const filteredUrls = data.iceServers.urls.filter((url: string) => !url.includes(':53'));

    this.creds = {
      username: data.iceServers.username,
      credential: data.iceServers.credential,
      urls: filteredUrls,
      expiresAt: now + (ttl * 1000) - 60000
    };

    return this.buildIceServers(this.creds);
  }

  private buildIceServers(c: { username: string; credential: string; urls: string[] }): RTCIceServer[] {
    return [
      { urls: 'stun:stun.cloudflare.com:3478' },
      { urls: c.urls, username: c.username, credential: c.credential, credentialType: 'password' as const }
    ];
  }
}
```

## Common Use Cases

```typescript
// Video conferencing: TURN as fallback
const config = { iceServers: await getTURNConfig(), iceTransportPolicy: 'all' };

// IoT/predictable connectivity: force TURN
const config = { iceServers: await getTURNConfig(), iceTransportPolicy: 'relay' };

// Screen sharing: reduce overhead
const pc = new RTCPeerConnection({ iceServers: await getTURNConfig(), bundlePolicy: 'max-bundle' });
```

## Integration with Cloudflare Calls SFU

```typescript
// TURN is automatically used when needed
// Cloudflare Calls handles TURN + SFU coordination
const session = await callsClient.createSession({
  appId: 'your-app-id',
  sessionId: 'meeting-123'
});
```

## Debugging ICE Connectivity

```typescript
pc.addEventListener('icecandidate', (event) => {
  if (event.candidate) {
    console.log('ICE candidate:', event.candidate.type, event.candidate.protocol);
  }
});

pc.addEventListener('iceconnectionstatechange', () => {
  console.log('ICE state:', pc.iceConnectionState);
});

// Check selected candidate pair
const stats = await pc.getStats();
stats.forEach(report => {
  if (report.type === 'candidate-pair' && report.selected) {
    console.log('Selected:', report);
  }
});
```

## See Also

- [api.md](./api.md) - Credential generation API, types
- [configuration.md](./configuration.md) - Worker setup, environment variables
- [gotchas.md](./gotchas.md) - Common mistakes, troubleshooting

# Cloudflare TURN Service

Expert guidance for implementing Cloudflare TURN Service in WebRTC applications.

## Overview

Cloudflare TURN (Traversal Using Relays around NAT) Service is a managed relay service for WebRTC applications. TURN acts as a relay point for traffic between WebRTC clients and SFUs, particularly when direct peer-to-peer communication is obstructed by NATs or firewalls. The service runs on Cloudflare's global anycast network across 310+ cities.

## Key Characteristics

- **Anycast Architecture**: Automatically connects clients to the closest Cloudflare location
- **Global Network**: Available across Cloudflare's entire network (excluding China Network)
- **Zero Configuration**: No need to manually select regions or servers
- **Protocol Support**: STUN/TURN over UDP, TCP, and TLS
- **Free Tier**: Free when used with Cloudflare Calls SFU, otherwise $0.05/GB outbound

## In This Reference

| File | Purpose |
|------|---------|
| [api.md](./api.md) | Credentials API, TURN key management, types, constraints |
| [configuration.md](./configuration.md) | Worker setup, wrangler.jsonc, env vars, IP allowlisting |
| [patterns.md](./patterns.md) | Implementation patterns, use cases, integration examples |
| [gotchas.md](./gotchas.md) | Troubleshooting, limits, security, common mistakes |

## Reading Order

| Task | Files to Read | Est. Tokens |
|------|---------------|-------------|
| Quick start | README only | ~500 |
| Generate credentials | README → api | ~1300 |
| Worker integration | README → configuration → patterns | ~2000 |
| Debug connection | gotchas | ~700 |
| Security review | api → gotchas | ~1500 |
| Enterprise firewall | configuration | ~600 |

## Service Addresses and Ports

### STUN over UDP
- **Primary**: `stun.cloudflare.com:3478/udp`
- **Alternate**: `stun.cloudflare.com:53/udp` (blocked by browsers, not recommended)

### TURN over UDP
- **Primary**: `turn.cloudflare.com:3478/udp`
- **Alternate**: `turn.cloudflare.com:53/udp` (blocked by browsers)

### TURN over TCP
- **Primary**: `turn.cloudflare.com:3478/tcp`
- **Alternate**: `turn.cloudflare.com:80/tcp`

### TURN over TLS
- **Primary**: `turn.cloudflare.com:5349/tcp`
- **Alternate**: `turn.cloudflare.com:443/tcp`

## Quick Start

1. **Create TURN key via API**: see [api.md#create-turn-key](./api.md#create-turn-key)
2. **Generate credentials**: see [api.md#generate-temporary-credentials](./api.md#generate-temporary-credentials)
3. **Configure Worker**: see [configuration.md#cloudflare-worker-integration](./configuration.md#cloudflare-worker-integration)
4. **Implement client**: see [patterns.md#basic-turn-configuration-browser](./patterns.md#basic-turn-configuration-browser)

## When to Use TURN

- **Restrictive NATs**: Symmetric NATs that block direct connections
- **Corporate firewalls**: Environments blocking WebRTC ports
- **Mobile networks**: Carrier-grade NAT scenarios
- **Predictable connectivity**: When reliability > efficiency

## Related Cloudflare Services

- **Cloudflare Calls SFU**: Managed Selective Forwarding Unit (TURN free when used with SFU)
- **Cloudflare Stream**: Video streaming with WHIP/WHEP support
- **Cloudflare Workers**: Backend for credential generation
- **Cloudflare KV**: Credential caching
- **Cloudflare Durable Objects**: Session state management

## Additional Resources

- [Cloudflare Calls Documentation](https://developers.cloudflare.com/calls/)
- [Cloudflare TURN Service Docs](https://developers.cloudflare.com/realtime/turn/)
- [Cloudflare API Reference](https://developers.cloudflare.com/api/resources/calls/subresources/turn/)
- [Orange Meets (Open Source Example)](https://github.com/cloudflare/orange)

# Vectorize API routes

Fetch the relevant section of the [Workers binding API](https://developers.cloudflare.com/vectorize/reference/client-api/) before writing calls or types.

| Task | Current documentation |
|------|-----------------------|
| Vector shape, binding, and generated TypeScript types | [Vectorize API](https://developers.cloudflare.com/vectorize/reference/client-api/) |
| Insert, upsert, retrieve by ID, delete, or inspect an index | [Operations](https://developers.cloudflare.com/vectorize/reference/client-api/#operations) |
| Query by vector or ID; choose returned metadata, values, and scoring precision | [Query vectors](https://developers.cloudflare.com/vectorize/best-practices/query-vectors/) and [query options](https://developers.cloudflare.com/vectorize/reference/client-api/#query-vectors) |
| Filter by metadata, combine conditions, or use nested properties | [Metadata filtering](https://developers.cloudflare.com/vectorize/reference/metadata-filtering/) |
| Batch ingestion and select vector formats | [Insert vectors](https://developers.cloudflare.com/vectorize/best-practices/insert-vectors/) and [current limits](https://developers.cloudflare.com/vectorize/platform/limits/) |
| Manage indexes or vectors outside a Worker | [Wrangler commands](https://developers.cloudflare.com/vectorize/reference/wrangler-commands/) and [REST API](https://developers.cloudflare.com/api/resources/vectorize/subresources/indexes/methods/list/) |

## Operation choices

- Choose insert when existing IDs should be preserved; choose upsert when they should be replaced. Upsert replaces the whole vector, including metadata, so provide the complete intended record.
- Request only the values and metadata the caller needs. Indexed metadata can omit fields or truncate strings; full metadata and vector values change query limits and latency. Fetch the current query options before choosing a result count.
- Treat accepted mutations and query visibility as separate events. Use current mutation guidance when implementing ingestion verification or read-after-write behavior.

# Vectorize configuration routes

| Task | Current documentation |
|------|-----------------------|
| Create an index, choose dimensions and metric | [Create indexes](https://developers.cloudflare.com/vectorize/best-practices/create-indexes/) |
| Bind an index to a Worker, develop, deploy, and verify queries | [Introduction to Vectorize](https://developers.cloudflare.com/vectorize/get-started/intro/) |
| Configure bindings and generate types | [Binding and TypeScript guidance](https://developers.cloudflare.com/vectorize/reference/client-api/#binding-to-a-worker) |
| Create, list, or delete metadata indexes | [Metadata filtering](https://developers.cloudflare.com/vectorize/reference/metadata-filtering/) and [Wrangler commands](https://developers.cloudflare.com/vectorize/reference/wrangler-commands/) |
| Manage indexes and vectors through the CLI | [Wrangler commands](https://developers.cloudflare.com/vectorize/reference/wrangler-commands/) |
| Upload NDJSON and batch ingestion | [Insert vectors](https://developers.cloudflare.com/vectorize/best-practices/insert-vectors/) |
| Check capacity, payload, namespace, or batch constraints | [Limits](https://developers.cloudflare.com/vectorize/platform/limits/) |

## Configuration decisions

Confirm the embedding model, output dimensions, and distance metric before provisioning: dimensions and metric cannot be changed in place. Plan a new index and re-embedding where needed when changing models.

Create metadata indexes before ingesting vectors that must be filterable. If adding one to an existing dataset, plan to re-upsert the affected vectors after index creation.

Choose metadata granularity around actual queries. For range filters over high-cardinality fields, consider buckets that preserve the application's required precision; do not bucket identifiers used for exact matches. Fetch the [cardinality guidance](https://developers.cloudflare.com/vectorize/best-practices/insert-vectors/#performance-tips-when-filtering-by-metadata) before designing the schema.

# Vectorize troubleshooting routes

Fetch current documentation before diagnosing a numeric limit, API error, or delayed mutation. Do not infer batch sizes or result limits from old snippets.

| Symptom or decision | What to check | Current documentation |
|---------------------|---------------|-----------------------|
| A write succeeded but search has not changed | Mutations are asynchronous; acceptance does not guarantee query visibility | [Insert, upsert, and delete semantics](https://developers.cloudflare.com/vectorize/reference/client-api/#operations) |
| Ingestion is slow or a batch is rejected | Batch size depends on the interface; inspect throughput and payload constraints | [Write throughput](https://developers.cloudflare.com/vectorize/best-practices/insert-vectors/#improve-write-throughput) and [limits](https://developers.cloudflare.com/vectorize/platform/limits/) |
| Query count is rejected or metadata is incomplete | Returned values and metadata affect query limits; indexed metadata can be truncated | [Query options](https://developers.cloudflare.com/vectorize/reference/client-api/#query-vectors) |
| Metadata filters return no matches | Confirm field type, operators, nesting, and index creation; re-upsert data written before the metadata index existed | [Metadata filtering](https://developers.cloudflare.com/vectorize/reference/metadata-filtering/) |
| Query has no matches or poor relevance | Check embedding model and dimensions, metric, namespace, filters, and mutation visibility | [Query vectors](https://developers.cloudflare.com/vectorize/best-practices/query-vectors/) and [index configuration](https://developers.cloudflare.com/vectorize/best-practices/create-indexes/) |
| Existing IDs or metadata behave unexpectedly on update | Insert preserves existing IDs; upsert replaces the full vector and metadata | [Mutation semantics](https://developers.cloudflare.com/vectorize/reference/client-api/#operations) |
| Capacity or model output no longer fits | Check current limits and model output dimensions; changing dimensions or metric requires another index | [Limits](https://developers.cloudflare.com/vectorize/platform/limits/) and [create indexes](https://developers.cloudflare.com/vectorize/best-practices/create-indexes/) |

For changes to embedding providers or tenant boundaries, also read [pattern decisions](patterns.md).

# Vectorize pattern routes

| Task | Current documentation |
|------|-----------------------|
| Generate and query Workers AI embeddings | [Vectorize and Workers AI](https://developers.cloudflare.com/vectorize/get-started/embeddings/) |
| Query with embeddings from OpenAI | [OpenAI integration](https://developers.cloudflare.com/vectorize/best-practices/query-vectors/#openai) |
| Choose embedding dimensions and distance metric | [Create indexes](https://developers.cloudflare.com/vectorize/best-practices/create-indexes/) |
| Build a retrieval-augmented generation application | [Workers AI RAG tutorial](https://developers.cloudflare.com/workers-ai/guides/tutorials/build-a-retrieval-augmented-generation-ai/) |
| Link search results to source documents | [Vector metadata](https://developers.cloudflare.com/vectorize/best-practices/insert-vectors/#metadata) |
| Partition vectors by tenant | [Namespaces](https://developers.cloudflare.com/vectorize/best-practices/insert-vectors/#namespaces) and [namespace versus metadata filtering](https://developers.cloudflare.com/vectorize/reference/metadata-filtering/#namespace-versus-metadata-filtering) |
| Combine similarity search with categorical or range filters | [Metadata filtering](https://developers.cloudflare.com/vectorize/reference/metadata-filtering/) |
| Ingest or update vectors in batches | [Insert vectors](https://developers.cloudflare.com/vectorize/best-practices/insert-vectors/) and [limits](https://developers.cloudflare.com/vectorize/platform/limits/) |

## Embedding and retrieval decisions

Keep ingestion and query embeddings compatible: use the same model and preprocessing, and extract the individual vector from the provider's documented response shape. Fetch the selected model's current documentation for dimensions and input requirements.

For RAG, store a reliable reference to the source content and request the metadata needed to resolve it. Handle missing or deleted source documents before passing retrieved context to generation.

## Tenant scope

Namespaces and metadata filters narrow searches; they do not authenticate the caller. Derive the permitted tenant scope from trusted identity and enforce it on every relevant read and write, including ID-based retrieval and deletion. Do not assume a namespace query option protects other operations.

Choose namespace or metadata partitioning based on the required query scope and current limits. Both narrow the search space; avoid assuming metadata filtering happens after vector search. If tenant IDs are stored in metadata, create the corresponding metadata index before ingestion.

# Cloudflare Vectorize

Use Vectorize when you need to control embeddings, vector indexing, and retrieval for semantic search, recommendations, or RAG. For a managed retrieval pipeline, see [AI Search](../ai-search/README.md).

Fetch current documentation before implementing. Start with the [Vectorize documentation index](https://developers.cloudflare.com/vectorize/llms.txt) to discover pages; load only those relevant to the task. Treat the docs as the source of truth for APIs, configuration, models, limits, and pricing.

## Task routing

| Task | Read |
|------|------|
| Create an index and connect a Worker | [Configuration](configuration.md) and [Introduction to Vectorize](https://developers.cloudflare.com/vectorize/get-started/intro/) |
| Insert, update, query, retrieve, or delete vectors | [API routes](api.md) |
| Generate embeddings, build RAG, or partition tenant data | [Patterns](patterns.md) |
| Diagnose missing matches, metadata, or rejected requests | [Gotchas](gotchas.md) |

## Decisions to make first

- Use a consistent embedding model and preprocessing for stored vectors and queries. Matching dimensions alone does not make different models' embeddings compatible.
- Choose dimensions from the embedding output and a distance metric appropriate to that model. Changing either requires a new index; check [index configuration and scoring semantics](https://developers.cloudflare.com/vectorize/best-practices/create-indexes/) before choosing thresholds.
- Plan filterable metadata before ingestion. Adding an index later requires re-upserting existing vectors to index that metadata; see [metadata filtering](https://developers.cloudflare.com/vectorize/reference/metadata-filtering/).
- A namespace partitions search; your application must authorize access and derive tenant scope from trusted identity. See [tenant patterns](patterns.md).
- Design for asynchronous mutation visibility rather than assuming a completed write is already searchable. See [mutation semantics](https://developers.cloudflare.com/vectorize/reference/client-api/).

# WAF API Reference

Read the matching workflow before writing API calls or translating them into the project's installed SDK. Retrieve identifiers from the target account or zone; do not reuse example IDs.

| Task | Documentation |
|------|---------------|
| Inspect the entry point and add custom rules | [Create a custom rule via API](https://developers.cloudflare.com/waf/custom-rules/create-api/) |
| Discover managed rulesets and deploy them | [Deploy managed rules via API](https://developers.cloudflare.com/waf/managed-rules/deploy-api/) |
| Create rate limits with the current request schema | [Create a rate limiting rule via API](https://developers.cloudflare.com/waf/rate-limiting-rules/create-api/) |
| Select counting characteristics, expressions, periods, and mitigation behavior | [Rate limiting parameters](https://developers.cloudflare.com/waf/rate-limiting-rules/parameters/) |
| Replace a ruleset deliberately or choose an additive operation | [Update or deploy a ruleset](https://developers.cloudflare.com/ruleset-engine/rulesets-api/update/) |
| Construct expressions using supported fields, operators, and functions | [Rules language](https://developers.cloudflare.com/ruleset-engine/rules-language/) |
| Choose actions and understand terminating behavior | [Actions reference](https://developers.cloudflare.com/ruleset-engine/rules-language/actions/) |
| Override managed rules, tags, or a ruleset | [Managed ruleset overrides](https://developers.cloudflare.com/ruleset-engine/managed-rulesets/override-managed-ruleset/) |
| Choose exactly what a skip rule bypasses | [Skip options](https://developers.cloudflare.com/waf/custom-rules/skip/options/) |

For an addition, prefer the workflow's operation that adds a rule to an existing ruleset. When replacing a ruleset, include every rule that must remain; review the resulting rule list before applying it. See [gotchas.md](gotchas.md) for scope and evaluation checks.

# WAF Configuration

Identify the account or zone, existing rulesets, and the system managing them before choosing a deployment method. Use the permission requirements in the selected workflow rather than a copied token-permission list.

| Task | Documentation |
|------|---------------|
| Enable protections and configure them in the dashboard | [WAF get started](https://developers.cloudflare.com/waf/get-started/) |
| Configure custom rules through the API or SDK | [Custom rules API workflow](https://developers.cloudflare.com/waf/custom-rules/create-api/) |
| Configure managed rules through the API or SDK | [Managed rules API workflow](https://developers.cloudflare.com/waf/managed-rules/deploy-api/) |
| Configure rate limiting through the API or SDK | [Rate limiting API workflow](https://developers.cloudflare.com/waf/rate-limiting-rules/create-api/) |
| Manage custom rules with Terraform | [Custom rules Terraform guide](https://developers.cloudflare.com/terraform/additional-configurations/waf-custom-rules/) |
| Manage managed rulesets with Terraform | [Managed rules Terraform guide](https://developers.cloudflare.com/terraform/additional-configurations/waf-managed-rulesets/) |
| Manage rate limits with Terraform | [Rate limiting Terraform guide](https://developers.cloudflare.com/terraform/additional-configurations/rate-limiting-rules/) |
| Decide account versus zone placement and phase | [WAF phases](https://developers.cloudflare.com/waf/reference/phases/) |

When adopting existing rules into Terraform, follow the selected guide's import instructions and inspect the plan for unintended removals. Keep the existing management tool when it fits the task. For SDK or Pulumi projects, verify the installed package's types before translating the documented API workflow.

Continue with [patterns.md](patterns.md) for protection choices and [gotchas.md](gotchas.md) for diagnostics.

# WAF Gotchas

Use observed requests and the deployed ruleset definitions to diagnose behavior before changing protection.

| Symptom or decision | Documentation to read |
|---------------------|-----------------------|
| A rule executes earlier or later than expected | [WAF phases and account/zone order](https://developers.cloudflare.com/waf/reference/phases/) and [terminating actions](https://developers.cloudflare.com/ruleset-engine/rules-language/actions/) |
| A skip rule leaves a protection active | [Skip scope, phases, products, and logging](https://developers.cloudflare.com/waf/custom-rules/skip/options/) |
| Updating a ruleset removes unrelated rules | [Ruleset replacement semantics](https://developers.cloudflare.com/ruleset-engine/rulesets-api/update/) |
| An expression fails to parse | [Rules language elements](https://developers.cloudflare.com/ruleset-engine/rules-language/) |
| Score-based rules match unexpected traffic | [Attack score meaning, special values, and plan availability](https://developers.cloudflare.com/waf/detections/attack-score/) |
| Managed overrides conflict | [Override precedence and scope](https://developers.cloudflare.com/ruleset-engine/managed-rulesets/override-managed-ruleset/) |
| Legitimate traffic is blocked, or attacks reach the origin | [False-positive and false-negative investigation](https://developers.cloudflare.com/waf/managed-rules/troubleshooting/) |
| Rate limits affect shared-IP users or count unexpected requests | [Characteristics, NAT support, and counting expressions](https://developers.cloudflare.com/waf/rate-limiting-rules/parameters/) |
| API creation fails or the request body is unclear | [Custom rule creation](https://developers.cloudflare.com/waf/custom-rules/create-api/), [managed deployment](https://developers.cloudflare.com/waf/managed-rules/deploy-api/), or [rate limit creation](https://developers.cloudflare.com/waf/rate-limiting-rules/create-api/) |

Check account versus zone scope, rule position, and the action that actually handled the request. For rate limiting, inspect matching and counting criteria separately. Avoid broad exceptions as a shortcut for diagnosing a single false positive.

Return to [api.md](api.md) for operations and [configuration.md](configuration.md) for deployment ownership.

# WAF Patterns

Choose the workflow that matches the requested outcome, then retrieve its current examples. Make rule scope explicit and review the affected traffic before enforcement.

| Outcome | Documentation |
|---------|---------------|
| Apply managed protection to a zone | [Deploy managed rules](https://developers.cloudflare.com/waf/managed-rules/deploy-api/) |
| Change a managed rule's behavior or evaluate it with logging | [Override a managed ruleset](https://developers.cloudflare.com/ruleset-engine/managed-rulesets/override-managed-ruleset/) |
| Enforce an application-specific request policy | [Create a custom rule](https://developers.cloudflare.com/waf/custom-rules/create-api/) |
| Use attack detection in a request policy | [Attack score semantics and availability](https://developers.cloudflare.com/waf/detections/attack-score/) |
| Protect a login or API endpoint from excessive requests | [Rate limiting API examples](https://developers.cloudflare.com/waf/rate-limiting-rules/create-api/) and [counting parameters](https://developers.cloudflare.com/waf/rate-limiting-rules/parameters/) |
| Exempt narrowly identified traffic from selected protections | [Available skip options](https://developers.cloudflare.com/waf/custom-rules/skip/options/) |
| Adjust protection after a false positive | [Managed rules troubleshooting](https://developers.cloudflare.com/waf/managed-rules/troubleshooting/) |

For a false positive, identify the matching rule and request scope before choosing an exception or override. Keep the adjustment as narrow as the evidence supports. For combined protections, check [WAF phases](https://developers.cloudflare.com/waf/reference/phases/) before deciding where an exception belongs.

Use [configuration.md](configuration.md) to select the deployment method and [api.md](api.md) to preserve existing rules while changing it.

# Cloudflare WAF

Use this reference for managed protection, custom request policies, rate limiting, and investigation of blocked traffic. Read the relevant developer documentation before implementing; it owns schemas, expressions, ruleset IDs, phase order, and plan availability.

| Task | Start here |
|------|------------|
| Choose and enable WAF protections | [Get started](https://developers.cloudflare.com/waf/get-started/) |
| Deploy managed protection | [Managed rules deployment](https://developers.cloudflare.com/waf/managed-rules/deploy-api/) |
| Match application-specific requests | [Custom rules](https://developers.cloudflare.com/waf/custom-rules/create-api/) |
| Limit request volume | [Rate limiting](https://developers.cloudflare.com/waf/rate-limiting-rules/create-api/) |
| Understand score-based detection | [Attack score](https://developers.cloudflare.com/waf/detections/attack-score/) |
| Diagnose blocked or unmitigated requests | [Managed rules troubleshooting](https://developers.cloudflare.com/waf/managed-rules/troubleshooting/) |

Identify the target account or zone and inspect existing rules before planning a change. Keep the requested traffic scope explicit, especially for exceptions and account-wide deployments.

## Reading Order

1. [configuration.md](configuration.md) — deployment method and existing configuration.
2. [api.md](api.md) — API workflows and expression references.
3. [patterns.md](patterns.md) — choose a protection or exception workflow.
4. [gotchas.md](gotchas.md) — diagnose ordering, scope, and false positives.

# Configuration

## Setup Methods

### Proxied Sites (Automatic)

Dashboard → Web Analytics → Add site → Select hostname → Done

| Injection Option | Description |
|------------------|-------------|
| Enable | Auto-inject for all visitors (default) |
| Enable, excluding EU | No injection for EU (GDPR) |
| Enable with manual snippet | You add beacon manually |
| Disable | Pause tracking |

**Fails if response has:** `Cache-Control: public, no-transform`

**CSP required:**
```
script-src https://static.cloudflareinsights.com https://cloudflareinsights.com;
```

### Non-Proxied Sites (Manual)

Dashboard → Web Analytics → Add site → Enter hostname → Copy snippet

```html
<script defer src='https://static.cloudflareinsights.com/beacon.min.js' 
        data-cf-beacon='{"token": "YOUR_TOKEN", "spa": true}'></script>
```

**Limits:** 10 non-proxied sites per account

## SPA Mode

**Enable `spa: true` for:** React Router, Next.js, Vue Router, Nuxt, SvelteKit, Angular

**Keep `spa: false` for:** Traditional multi-page apps, static sites, WordPress

**Hash routing (`#/path`) NOT supported** - use History API routing.

## Token Management

- Found in: Dashboard → Web Analytics → Manage site
- **Not secrets** - domain-locked, safe to expose in HTML
- Each site gets unique token

## Environment Config

```typescript
// Only load in production
if (process.env.NODE_ENV === 'production') {
  // Load beacon
}
```

Or use environment-specific tokens via env vars.

## Verify Installation

1. DevTools Network → filter `cloudflareinsights` → see `beacon.min.js` + data request
2. No CSP/CORS errors in console
3. Dashboard shows pageviews after 5-10 min delay

## Rules (Plan-dependent)

Configure in dashboard for:
- **Sample rate** - reduce collection % for high-traffic
- **Path-based** - different behavior per route
- **Host-based** - separate tracking per domain

## Data Retention

- 6 months rolling window
- 1-hour bucket granularity
- No raw export, dashboard only

# Web Analytics Gotchas

## Critical Issues

### SPA Navigation Not Tracked

**Symptom:** Only initial pageload counted  
**Fix:** Add `spa: true`:
```html
<script data-cf-beacon='{"token": "TOKEN", "spa": true}' ...></script>
```

### CSP Blocking Beacon

**Symptom:** Console error "Refused to load script"  
**Fix:** Allow both domains:
```
script-src 'self' https://static.cloudflareinsights.com https://cloudflareinsights.com;
```

### Hash-Based Routing Unsupported

**Symptom:** `#/path` URLs not tracked  
**Fix:** Migrate to History API (`BrowserRouter`, not `HashRouter`). No workaround for hash routing.

### No Data Appearing

**Causes & Fixes:**
1. **Delay** - Wait 5-15 minutes
2. **Wrong token** - Verify matches dashboard exactly
3. **Script blocked** - Check DevTools Network tab for beacon.min.js
4. **Domain mismatch** - Dashboard site must match actual URL

### Auto-Injection Fails

**Cause:** `Cache-Control: no-transform` header  
**Fix:** Remove `no-transform` or install beacon manually

### Duplicate Pageviews

**Cause:** Multiple beacon scripts  
**Fix:** Keep only one beacon per page

## Configuration Issues

| Issue | Fix |
|-------|-----|
| 10-site limit reached | Delete old sites or proxy through CF (unlimited) |
| Token not recognized | Use exact alphanumeric token from dashboard |

## Framework-Specific

### Next.js Hydration Warning

```tsx
<script suppressHydrationWarning ... />
```

### Gatsby Window Undefined

Use `gatsby-browser.js` to load client-side only.

## Limits

| Resource | Limit |
|----------|-------|
| Non-proxied sites | 10 |
| Proxied sites | Unlimited |
| Data retention | 6 months |
| Ingestion delay | 5-10 min |
| API access | None (dashboard only) |

## When NOT to Use Web Analytics

Use alternatives if you need:
- Custom event tracking
- Real-time data
- User-level tracking
- Conversion funnels
- Data export/API access

**Web Analytics excels at:** Core Web Vitals, basic traffic, privacy compliance, free unlimited pageviews.

# Framework Integration

**Web Analytics is dashboard-only** - no programmatic API. This covers beacon integration.

## Basic HTML

```html
<script defer src='https://static.cloudflareinsights.com/beacon.min.js' 
        data-cf-beacon='{"token": "YOUR_TOKEN", "spa": true}'></script>
```

Place before closing `</body>` tag.

## Framework Examples

| Framework | Location | Notes |
|-----------|----------|-------|
| React/Vite | `public/index.html` | Add `spa: true` |
| Next.js App Router | `app/layout.tsx` | Use `<Script strategy="afterInteractive">` |
| Next.js Pages | `pages/_document.tsx` | Use `<Script>` |
| Nuxt 3 | `app.vue` with `useHead()` | Or use plugin |
| Vue 3/Vite | `index.html` | Add `spa: true` |
| Gatsby | `gatsby-browser.js` | `onClientEntry` hook |
| SvelteKit | `src/app.html` | Before `</body>` |
| Astro | Layout component | Before `</body>` |
| Angular | `src/index.html` | Add `spa: true` |
| Docusaurus | `docusaurus.config.js` | In `scripts` array |

## Configuration

```json
{
  "token": "YOUR_TOKEN",
  "spa": true
}
```

**Use `spa: true` for:** React Router, Vue Router, Next.js, Nuxt, Gatsby, SvelteKit, Angular

**Use `spa: false` for:** Traditional server-rendered (PHP, Django, Rails, WordPress)

## CSP Headers

```
script-src 'self' https://static.cloudflareinsights.com;
connect-src 'self' https://cloudflareinsights.com;
```

## GDPR Consent

```typescript
// Load conditionally based on consent
if (localStorage.getItem('analytics-consent') === 'true') {
  const script = document.createElement('script');
  script.src = 'https://static.cloudflareinsights.com/beacon.min.js';
  script.defer = true;
  script.setAttribute('data-cf-beacon', '{"token": "YOUR_TOKEN", "spa": true}');
  document.body.appendChild(script);
}
```

# Web Analytics Patterns

## Core Web Vitals Debugging

Dashboard → Core Web Vitals → Click metric → Debug View shows top 5 problematic elements.

### LCP Fixes

```html
<!-- Priority hints -->
<img src="hero.jpg" loading="eager" fetchpriority="high" />
<link rel="preload" as="image" href="/hero.jpg" fetchpriority="high" />
```

### CLS Fixes

```css
/* Reserve space */
.ad-container { min-height: 250px; }
img { width: 400px; height: 300px; } /* Explicit dimensions */
```

### INP Fixes

```typescript
// Debounce expensive operations
const handleInput = debounce(search, 300);

// Yield to main thread
await task(); await new Promise(r => setTimeout(r, 0)); await task2();

// Move to Web Worker for heavy computation
```

| Metric | Good | Poor |
|--------|------|------|
| LCP | ≤2.5s | >4s |
| INP | ≤200ms | >500ms |
| CLS | ≤0.1 | >0.25 |

## GDPR Consent

```typescript
// Load beacon only after consent
const consent = localStorage.getItem('analytics-consent');
if (consent === 'accepted') {
  const script = document.createElement('script');
  script.src = 'https://static.cloudflareinsights.com/beacon.min.js';
  script.setAttribute('data-cf-beacon', '{"token": "TOKEN", "spa": true}');
  document.body.appendChild(script);
}
```

Alternative: Dashboard → "Enable, excluding visitor data in the EU"

## SPA Navigation

```html
<!-- REQUIRED for React/Vue/etc routing -->
<script data-cf-beacon='{"token": "TOKEN", "spa": true}' ...></script>
```

Without `spa: true`: only initial pageload tracked.

## Staging/Production Separation

```typescript
// Use env-specific tokens
const token = process.env.NEXT_PUBLIC_CF_ANALYTICS_TOKEN;
// .env.production: production token
// .env.staging: staging token (or empty to disable)
```

## Bot Filtering

Dashboard → Filters → "Exclude Bot Traffic"

Filters: Search crawlers, monitoring services, known bots.  
Not filtered: Headless browsers (Playwright/Puppeteer).

## Ad-Blocker Impact

~25-40% of users may block `cloudflareinsights.com`. No official workaround.
Dashboard shows minimum baseline; use server logs for complete picture.

## Limitations

- No UTM parameter tracking
- No webhooks/alerts/API
- No custom beacon domains
- Max 10 non-proxied sites

# Cloudflare Web Analytics

Privacy-first web analytics providing Core Web Vitals, traffic metrics, and user insights without compromising visitor privacy.

## Overview

Cloudflare Web Analytics provides:
- **Core Web Vitals** - LCP, FID, CLS, INP, TTFB monitoring
- **Page views & visits** - Traffic patterns without cookies
- **Referrers & paths** - Traffic sources and popular pages
- **Device & browser data** - User agent breakdown
- **Geographic data** - Country-level visitor distribution
- **Privacy-first** - No cookies, fingerprinting, or PII collection
- **Free** - No cost, unlimited pageviews

**Important:** Web Analytics is **dashboard-only**. No API exists for programmatic data access.

## Quick Start Decision Tree

```
Is your site proxied through Cloudflare?
├─ YES → Use automatic injection (configuration.md)
│   ├─ Enable auto-injection in dashboard
│   └─ No code changes needed (unless Cache-Control: no-transform)
│
└─ NO → Use manual beacon integration (integration.md)
    ├─ Add JS snippet to HTML
    ├─ Use spa: true for React/Vue/Next.js
    └─ Configure CSP if needed
```

## Reading Order

1. **[configuration.md](configuration.md)** - Setup for proxied vs non-proxied sites
2. **[integration.md](integration.md)** - Framework-specific beacon integration (React, Next.js, Vue, Nuxt, etc.)
3. **[patterns.md](patterns.md)** - Common use cases (performance monitoring, GDPR consent, multi-site tracking)
4. **[gotchas.md](gotchas.md)** - Troubleshooting (SPA tracking, CSP issues, hash routing limitations)

## When to Use Each File

- **Setting up for first time?** → Start with configuration.md
- **Using React/Next.js/Vue/Nuxt?** → Go to integration.md for framework code
- **Need GDPR consent loading?** → See patterns.md
- **Beacon not loading or no data?** → Check gotchas.md
- **SPA not tracking navigation?** → See integration.md for `spa: true` config

## Key Concepts

### Proxied vs Non-Proxied Sites

| Type | Description | Beacon Injection | Limit |
|------|-------------|------------------|-------|
| **Proxied** | DNS through Cloudflare (orange cloud) | Automatic or manual | Unlimited |
| **Non-proxied** | External hosting, manual beacon | Manual only | 10 sites max |

### SPA Mode

**Critical for modern frameworks:**
```json
{"token": "YOUR_TOKEN", "spa": true}
```

Without `spa: true`, client-side navigation (React Router, Vue Router, Next.js routing) will NOT be tracked. Only initial page loads will register.

### CSP Requirements

If using Content Security Policy, allow both domains:
```
script-src https://static.cloudflareinsights.com https://cloudflareinsights.com;
```

## Features

### Core Web Vitals Debugging
- **LCP (Largest Contentful Paint)** - Identifies slow-loading hero images/elements
- **FID (First Input Delay)** - Interaction responsiveness (legacy metric)
- **INP (Interaction to Next Paint)** - Modern interaction responsiveness metric
- **CLS (Cumulative Layout Shift)** - Visual stability issues
- **TTFB (Time to First Byte)** - Server response performance

Dashboard shows top 5 problematic elements with CSS selectors for debugging.

### Traffic Filters
- **Bot filtering** - Exclude automated traffic from metrics
- **Date ranges** - Custom time period analysis
- **Geographic** - Country-level filtering
- **Device type** - Desktop, mobile, tablet breakdown
- **Browser/OS** - User agent filtering

### Rules (Advanced - Plan-dependent)

Create custom tracking rules for advanced configurations:

**Sample Rate Rules:**
- Reduce data collection percentage for high-traffic sites
- Example: Track only 50% of visitors to reduce volume

**Path-Based Rules:**
- Different behavior per route
- Example: Exclude `/admin/*` or `/internal/*` from tracking

**Host-Based Rules:**
- Multi-domain configurations
- Example: Separate tracking for staging vs production subdomains

**Availability:** Rules feature depends on your Cloudflare plan. Check dashboard under Web Analytics → Rules to see if available. Free plans may have limited or no access.

## Plan Limits

| Feature | Free | Notes |
|---------|------|-------|
| Proxied sites | Unlimited | DNS through Cloudflare |
| Non-proxied sites | 10 | External hosting |
| Pageviews | Unlimited | No volume limits |
| Data retention | 6 months | Rolling window |
| Rules | Plan-dependent | Check dashboard |

## Privacy & Compliance

- **No cookies** - Zero client-side storage
- **No fingerprinting** - No tracking across sites
- **No PII** - IP addresses not stored
- **GDPR-friendly** - Minimal data collection
- **CCPA-compliant** - No personal data sale

**EU opt-out:** Dashboard option to exclude EU visitor data entirely.

## Limitations

- **Dashboard-only** - No API for programmatic access
- **No real-time** - 5-10 minute data delay
- **No custom events** - Automatic pageview/navigation tracking only
- **History API only** - Hash-based routing (`#/path`) not supported
- **No session replay** - Metrics only, no user recordings
- **No form tracking** - Page navigation tracking only

## See Also

- [Cloudflare Web Analytics Docs](https://developers.cloudflare.com/analytics/web-analytics/)
- [Core Web Vitals Guide](https://web.dev/vitals/)
- [GraphQL Analytics API Reference](../graphql-api/) - Query server-side analytics (HTTP, Workers, DNS, Firewall, etc.) via GraphQL

# Workerd APIs

## Worker Code (JS/TS)

### ES Modules (Recommended)
```javascript
export default {
  async fetch(request, env, ctx) {
    const value = await env.KV.get("key");           // Bindings in env
    const response = await env.API.fetch(request);   // Service binding
    ctx.waitUntil(logRequest(request));              // Background task
    return new Response("OK");
  },
  async adminApi(request, env, ctx) { /* Named entrypoint */ },
  async queue(batch, env, ctx) { /* Queue consumer */ },
  async scheduled(event, env, ctx) { /* Cron handler */ }
};
```

### TypeScript Types

**Generate from wrangler.toml (Recommended):**
```bash
wrangler types  # Output: worker-configuration.d.ts
```

**Manual types:**
```typescript
interface Env {
  API: Fetcher;
  CACHE: KVNamespace;
  STORAGE: R2Bucket;
  ROOMS: DurableObjectNamespace;
  API_KEY: string;
}

export default {
  async fetch(request: Request, env: Env, ctx: ExecutionContext): Promise<Response> {
    return new Response(await env.CACHE.get("key"));
  }
};
```

**Setup:**
```bash
npm install -D @cloudflare/workers-types
```

```json
// tsconfig.json
{"compilerOptions": {"types": ["@cloudflare/workers-types"]}}
```

### Service Worker Syntax (Legacy)
```javascript
addEventListener('fetch', event => {
  event.respondWith(handleRequest(event.request));
});

async function handleRequest(request) {
  const value = await KV.get("key");  // Bindings as globals
  return new Response("OK");
}
```

### Durable Objects
```javascript
export class Room {
  constructor(state, env) { this.state = state; this.env = env; }
  
  async fetch(request) {
    const url = new URL(request.url);
    if (url.pathname === "/increment") {
      const value = (await this.state.storage.get("counter")) || 0;
      await this.state.storage.put("counter", value + 1);
      return new Response(String(value + 1));
    }
    return new Response("Not found", {status: 404});
  }
}
```

### RPC Between Services
```javascript
// Caller: env.AUTH.validateToken(token) returns structured data
const user = await env.AUTH.validateToken(request.headers.get("Authorization"));

// Callee: export methods that return data
export default {
  async validateToken(token) { return {id: 123, name: "Alice"}; }
};
```

## Web Platform APIs

### Fetch
- `fetch()`, `Request`, `Response`, `Headers`
- `AbortController`, `AbortSignal`

### Streams
- `ReadableStream`, `WritableStream`, `TransformStream`
- Byte streams, BYOB readers

### Web Crypto
- `crypto.subtle` (encrypt/decrypt/sign/verify)
- `crypto.randomUUID()`, `crypto.getRandomValues()`

### Encoding
- `TextEncoder`, `TextDecoder`
- `atob()`, `btoa()`

### Web Standards
- `URL`, `URLSearchParams`
- `Blob`, `File`, `FormData`
- `WebSocket`

### Server-Sent Events (EventSource)
```javascript
// Server-side SSE
const { readable, writable } = new TransformStream();
const writer = writable.getWriter();
writer.write(new TextEncoder().encode('data: Hello\n\n'));
return new Response(readable, {headers: {'Content-Type': 'text/event-stream'}});
```

### HTMLRewriter (HTML Parsing/Transformation)
```javascript
const response = await fetch('https://example.com');
return new HTMLRewriter()
  .on('a[href]', {
    element(el) {
      el.setAttribute('href', `/proxy?url=${encodeURIComponent(el.getAttribute('href'))}`);
    }
  })
  .on('script', { element(el) { el.remove(); } })
  .transform(response);
```

### TCP Sockets (Experimental)
```javascript
const socket = await connect({ hostname: 'example.com', port: 80 });
const writer = socket.writable.getWriter();
await writer.write(new TextEncoder().encode('GET / HTTP/1.1\r\n\r\n'));
const reader = socket.readable.getReader();
const { value } = await reader.read();
return new Response(value);
```

### Performance
- `performance.now()`, `performance.timeOrigin`
- `setTimeout()`, `setInterval()`, `queueMicrotask()`

### Console
- `console.log()`, `console.error()`, `console.warn()`

### Node.js Compat (`nodejs_compat` flag)
```javascript
import { Buffer } from 'node:buffer';
import { randomBytes } from 'node:crypto';

const buf = Buffer.from('Hello');
const random = randomBytes(16);
```

**Available:** `node:buffer`, `node:crypto`, `node:stream`, `node:util`, `node:events`, `node:assert`, `node:path`, `node:querystring`, `node:url`
**NOT available:** `node:fs`, `node:http`, `node:net`, `node:child_process`

## CLI Commands

```bash
workerd serve config.capnp [constantName]          # Start server
workerd serve config.capnp --socket-addr http=*:3000 --verbose
workerd compile config.capnp constantName -o binary  # Compile to binary
workerd test config.capnp [--test-only=test.js]    # Run tests
```

## Wrangler Integration

Use Wrangler for development:
```bash
wrangler dev     # Uses workerd internally
wrangler types   # Generate TypeScript types from wrangler.toml
```

See [patterns.md](./patterns.md) for usage examples, [configuration.md](./configuration.md) for config details.

# Workerd Configuration

## Basic Structure
```capnp
using Workerd = import "/workerd/workerd.capnp";

const config :Workerd.Config = (
  services = [(name = "main", worker = .mainWorker)],
  sockets = [(name = "http", address = "*:8080", http = (), service = "main")]
);

const mainWorker :Workerd.Worker = (
  modules = [(name = "index.js", esModule = embed "src/index.js")],
  compatibilityDate = "2024-01-15",
  bindings = [...]
);
```

## Services
**Worker**: Run JS/Wasm code
```capnp
(name = "api", worker = (
  modules = [(name = "index.js", esModule = embed "index.js")],
  compatibilityDate = "2024-01-15",
  bindings = [...]
))
```

**Network**: Internet access
```capnp
(name = "internet", network = (allow = ["public"], tlsOptions = (trustBrowserCas = true)))
```

**External**: Reverse proxy
```capnp
(name = "backend", external = (address = "api.com:443", http = (style = tls)))
```

**Disk**: Static files
```capnp
(name = "assets", disk = (path = "/var/www", writable = false))
```

## Sockets
```capnp
(name = "http", address = "*:8080", http = (), service = "main")
(name = "https", address = "*:443", https = (options = (), tlsOptions = (keypair = (...))), service = "main")
(name = "app", address = "unix:/tmp/app.sock", http = (), service = "main")
```

## Worker Formats
```capnp
# ES Modules (recommended)
modules = [(name = "index.js", esModule = embed "src/index.js"), (name = "wasm.wasm", wasm = embed "build/module.wasm")]

# Service Worker (legacy)
serviceWorkerScript = embed "worker.js"

# CommonJS
(name = "legacy.js", commonJsModule = embed "legacy.js", namedExports = ["foo"])
```

## Bindings
Bindings expose resources to workers. ES modules: `env.BINDING`, Service workers: globals.

### Primitive Types
```capnp
(name = "API_KEY", text = "secret")                    # String
(name = "CONFIG", json = '{"key":"val"}')              # Parsed JSON
(name = "DATA", data = embed "data.bin")               # ArrayBuffer
(name = "DATABASE_URL", fromEnvironment = "DB_URL")    # System env var
```

### Service Binding
```capnp
(name = "AUTH", service = "auth-worker")               # Basic
(name = "API", service = (
  name = "backend",
  entrypoint = "adminApi",                             # Named export
  props = (json = '{"role":"admin"}')                  # ctx.props
))
```

### Storage
```capnp
(name = "CACHE", kvNamespace = "kv-service")           # KV
(name = "STORAGE", r2Bucket = "r2-service")            # R2
(name = "ROOMS", durableObjectNamespace = (
  serviceName = "room-service",
  className = "Room"
))
(name = "FAST", memoryCache = (
  id = "cache-id",
  limits = (maxKeys = 1000, maxValueSize = 1048576)
))
```

### Other
```capnp
(name = "TASKS", queue = "queue-service")
(name = "ANALYTICS", analyticsEngine = "analytics")
(name = "LOADER", workerLoader = (id = "dynamic"))
(name = "KEY", cryptoKey = (format = raw, algorithm = (name = "HMAC", hash = "SHA-256"), keyData = embed "key.bin", usages = [sign, verify], extractable = false))
(name = "TRACED", wrapped = (moduleName = "tracing", entrypoint = "makeTracer", innerBindings = [(name = "backend", service = "backend")]))
```

## Compatibility
```capnp
compatibilityDate = "2024-01-15"                       # Always set!
compatibilityFlags = ["nodejs_compat", "streams_enable_constructors"]
```

Version = max compat date. Update carefully after testing.

## Parameter Bindings (Inheritance)
```capnp
const base :Workerd.Worker = (
  modules = [...], compatibilityDate = "2024-01-15",
  bindings = [(name = "API_URL", parameter = (type = text)), (name = "DB", parameter = (type = service))]
);

const derived :Workerd.Worker = (
  inherit = "base-service",
  bindings = [(name = "API_URL", text = "https://api.com"), (name = "DB", service = "postgres")]
);
```

## Durable Objects Config
```capnp
const worker :Workerd.Worker = (
  modules = [...],
  compatibilityDate = "2024-01-15",
  bindings = [(name = "ROOMS", durableObjectNamespace = "Room")],
  durableObjectNamespaces = [(className = "Room", uniqueKey = "v1")],
  durableObjectStorage = (localDisk = "/var/do")
);
```

## Remote Bindings (Development)

Connect local workerd to production Cloudflare resources:

```capnp
bindings = [
  # Remote KV (requires API token)
  (name = "PROD_KV", kvNamespace = (
    remote = (
      accountId = "your-account-id",
      namespaceId = "your-namespace-id",
      apiToken = .envVar("CF_API_TOKEN")
    )
  )),
  
  # Remote R2
  (name = "PROD_R2", r2Bucket = (
    remote = (
      accountId = "your-account-id",
      bucketName = "my-bucket",
      apiToken = .envVar("CF_API_TOKEN")
    )
  )),
  
  # Remote Durable Object
  (name = "PROD_DO", durableObjectNamespace = (
    remote = (
      accountId = "your-account-id",
      scriptName = "my-worker",
      className = "MyDO",
      apiToken = .envVar("CF_API_TOKEN")
    )
  ))
]
```

**Note:** Remote bindings require network access and valid Cloudflare API credentials.

## Logging & Debugging
```capnp
logging = (structuredLogging = true, stdoutPrefix = "OUT: ", stderrPrefix = "ERR: ")
v8Flags = ["--expose-gc", "--max-old-space-size=2048"]  # ⚠️ Unsupported in production
```

See [patterns.md](./patterns.md) for multi-service examples, [gotchas.md](./gotchas.md) for config errors.

# Workerd Gotchas

## Common Errors

### "Missing compatibility date"
**Cause:** Compatibility date not set
**Solution:**
❌ Wrong:
```capnp
const worker :Workerd.Worker = (
  serviceWorkerScript = embed "worker.js"
)
```

✅ Correct:
```capnp
const worker :Workerd.Worker = (
  serviceWorkerScript = embed "worker.js",
  compatibilityDate = "2024-01-15"  # Always set!
)
```

### Wrong Binding Type
**Problem:** JSON not parsed
**Cause:** Using `text = '{"key":"value"}'` instead of `json`
**Solution:** Use `json = '{"key":"value"}'` for parsed objects

### Service vs Namespace
**Problem:** Cannot create DO instance
**Cause:** Using `service = "room-service"` for Durable Object
**Solution:** Use `durableObjectNamespace = "Room"` for DO bindings

### Module Name Mismatch
**Problem:** Import fails
**Cause:** Module name includes path: `name = "src/index.js"`
**Solution:** Use simple names: `name = "index.js"`, embed with path

## Network Access

**Problem:** Fetch fails with network error
**Cause:** No network service configured (workerd has no global fetch)
**Solution:** Add network service binding:
```capnp
services = [(name = "internet", network = (allow = ["public"]))]
bindings = [(name = "NET", service = "internet")]
```

Or external service:
```capnp
bindings = [(name = "API", service = (external = (address = "api.com:443", http = (style = tls))))]
```

### "Worker not responding"
**Cause:** Socket misconfigured, no fetch handler, or port unavailable
**Solution:** Verify socket `address` matches, worker exports `fetch()`, port available

### "Binding not found"
**Cause:** Name mismatch or service doesn't exist
**Solution:** Check binding name in config matches code (`env.BINDING` for ES modules)

### "Module not found"
**Cause:** Module name doesn't match import or bad embed path
**Solution:** Module `name` must match import path exactly, verify `embed` path

### "Compatibility error"
**Cause:** Date not set or API unavailable on that date
**Solution:** Set `compatibilityDate`, verify API available on that date

## Performance Issues

**Problem:** High memory usage
**Cause:** Large caches or many isolates
**Solution:** Set cache limits, reduce isolate count, or use V8 flags (caution)

**Problem:** Slow startup
**Cause:** Many modules or complex config
**Solution:** Compile to binary (`workerd compile`), reduce imports

**Problem:** Request timeouts
**Cause:** External service issues or DNS problems
**Solution:** Check connectivity, DNS resolution, TLS handshake

## Build Issues

**Problem:** Cap'n Proto syntax errors
**Cause:** Invalid config or missing schema
**Solution:** Install capnproto tools, validate: `capnp compile -I. config.capnp`

**Problem:** Embed path not found
**Cause:** Path relative to config file
**Solution:** Use correct relative path or absolute path

**Problem:** V8 flags cause crashes
**Cause:** Unsafe V8 flags
**Solution:** ⚠️ V8 flags unsupported in production. Test thoroughly before use.

## Security Issues

**Problem:** Hardcoded secrets in config
**Cause:** `text` binding with secret value
**Solution:** Use `fromEnvironment` to load from env vars

**Problem:** Overly broad network access
**Cause:** `network = (allow = ["*"])`
**Solution:** Restrict to `allow = ["public"]` or specific hosts

**Problem:** Extractable crypto keys
**Cause:** `cryptoKey = (extractable = true, ...)`
**Solution:** Set `extractable = false` unless export required

## Compatibility Changes

**Problem:** Breaking changes after compat date update
**Cause:** New flags enabled between dates
**Solution:** Review [compat dates docs](https://developers.cloudflare.com/workers/configuration/compatibility-dates/), test locally first

**Problem:** "Compatibility date not supported"
**Cause:** Workerd version older than compat date
**Solution:** Update workerd binary (version = max compat date supported)

## Limits

| Resource/Limit | Value | Notes |
|----------------|-------|-------|
| V8 flags | Unsupported in production | Use with caution |
| Compatibility date | Must match workerd version | Update if mismatch |
| Module count | Affects startup time | Many imports slow |

## Troubleshooting Steps

1. **Enable verbose logging**: `workerd serve config.capnp --verbose`
2. **Check logs**: Look for error messages, stack traces
3. **Validate config**: `capnp compile -I. config.capnp`
4. **Test bindings**: Log `Object.keys(env)` to verify
5. **Check versions**: Workerd version vs compat date
6. **Isolate issue**: Minimal repro config
7. **Review schema**: [workerd.capnp](https://github.com/cloudflare/workerd/blob/main/src/workerd/server/workerd.capnp)

See [configuration.md](./configuration.md) for config details, [patterns.md](./patterns.md) for working examples, [api.md](./api.md) for runtime APIs.

# Workerd Patterns

## Multi-Service Architecture
```capnp
const config :Workerd.Config = (
  services = [
    (name = "frontend", worker = (
      modules = [(name = "index.js", esModule = embed "frontend/index.js")],
      compatibilityDate = "2024-01-15",
      bindings = [(name = "API", service = "api")]
    )),
    (name = "api", worker = (
      modules = [(name = "index.js", esModule = embed "api/index.js")],
      compatibilityDate = "2024-01-15",
      bindings = [(name = "DB", service = "postgres"), (name = "CACHE", kvNamespace = "kv")]
    )),
    (name = "postgres", external = (address = "db.internal:5432", http = ())),
    (name = "kv", disk = (path = "/var/kv", writable = true))
  ],
  sockets = [(name = "http", address = "*:8080", http = (), service = "frontend")]
);
```

## Durable Objects
```capnp
const worker :Workerd.Worker = (
  modules = [(name = "index.js", esModule = embed "index.js"), (name = "room.js", esModule = embed "room.js")],
  compatibilityDate = "2024-01-15",
  bindings = [(name = "ROOMS", durableObjectNamespace = "Room")],
  durableObjectNamespaces = [(className = "Room", uniqueKey = "v1")],
  durableObjectStorage = (localDisk = "/var/do")
);
```

## Dev vs Prod Configs
```capnp
# Use parameter bindings for env-specific config
const baseWorker :Workerd.Worker = (
  modules = [(name = "index.js", esModule = embed "src/index.js")],
  compatibilityDate = "2024-01-15",
  bindings = [(name = "API_URL", parameter = (type = text))]
);

const prodWorker :Workerd.Worker = (
  inherit = "base-service",
  bindings = [(name = "API_URL", text = "https://api.prod.com")]
);
```

## HTTP Reverse Proxy
```capnp
services = [
  (name = "proxy", worker = (serviceWorkerScript = embed "proxy.js", compatibilityDate = "2024-01-15", bindings = [(name = "BACKEND", service = "backend")])),
  (name = "backend", external = (address = "internal:8080", http = ()))
]
```

## Local Development

**Recommended:** Use Wrangler
```bash
wrangler dev  # Uses workerd internally
```

**Direct workerd:**
```bash
workerd serve config.capnp --socket-addr http=*:3000 --verbose
```

**Environment variables:**
```capnp
bindings = [(name = "DATABASE_URL", fromEnvironment = "DATABASE_URL")]
```

## Testing
```bash
workerd test config.capnp
workerd test config.capnp --test-only=test.js
```

Test files must be included in `modules = [...]` config.

## Production Deployment

### Compiled Binary (Recommended)
```bash
workerd compile config.capnp myConfig -o production-server
./production-server
```

### Docker
```dockerfile
FROM debian:bookworm-slim
RUN apt-get update && apt-get install -y ca-certificates
COPY workerd /usr/local/bin/
COPY config.capnp /etc/workerd/
COPY src/ /etc/workerd/src/
EXPOSE 8080
CMD ["workerd", "serve", "/etc/workerd/config.capnp"]
```

### Systemd
```ini
# /etc/systemd/system/workerd.service
[Service]
ExecStart=/usr/bin/workerd serve /etc/workerd/config.capnp --socket-fd http=3
Restart=always
User=nobody
```

See systemd socket activation docs for complete setup.

## Framework Integration

### Hono
```javascript
import { Hono } from 'hono';

const app = new Hono();

app.get('/', (c) => c.text('Hello Hono!'));
app.get('/api/:id', async (c) => {
  const id = c.req.param('id');
  const data = await c.env.KV.get(id);
  return c.json({ id, data });
});

export default app;
```

### itty-router
```javascript
import { Router } from 'itty-router';

const router = Router();

router.get('/', () => new Response('Hello itty!'));
router.get('/api/:id', async (request, env) => {
  const { id } = request.params;
  const data = await env.KV.get(id);
  return Response.json({ id, data });
});

export default {
  fetch: (request, env, ctx) => router.handle(request, env, ctx)
};
```

## Best Practices

1. **Use ES modules** over service worker syntax
2. **Explicit bindings** - no global namespace assumptions
3. **Type safety** - define `Env` interfaces (use `wrangler types`)
4. **Service isolation** - split concerns into multiple services
5. **Pin compat date** in production after testing
6. **Use ctx.waitUntil()** for background tasks
7. **Handle errors gracefully** with try/catch
8. **Configure resource limits** on caches/storage

## Common Patterns

### Error Handling
```javascript
export default {
  async fetch(request, env, ctx) {
    try {
      return await handleRequest(request, env);
    } catch (error) {
      console.error("Request failed", error);
      return new Response("Internal Error", {status: 500});
    }
  }
};
```

### Background Tasks
```javascript
export default {
  async fetch(request, env, ctx) {
    const response = new Response("OK");
    
    // Fire-and-forget background work
    ctx.waitUntil(
      env.ANALYTICS.put(request.url, Date.now())
    );
    
    return response;
  }
};
```

See [configuration.md](./configuration.md) for config syntax, [api.md](./api.md) for runtime APIs, [gotchas.md](./gotchas.md) for common errors.

# Workerd Runtime

V8-based JS/Wasm runtime powering Cloudflare Workers. Use as app server, dev tool, or HTTP proxy.

## ⚠️ IMPORTANT SECURITY NOTICE
**workerd is NOT a hardened sandbox.** Do not run untrusted code. It's designed for deploying YOUR code locally/self-hosted, not multi-tenant SaaS. Cloudflare production adds security layers not present in open-source workerd.

## Decision Tree: When to Use What

**95% of users:** Use Wrangler
- Local development: `wrangler dev` (uses workerd internally)
- Deployment: `wrangler deploy` (deploys to Cloudflare)
- Types: `wrangler types` (generates TypeScript types)

**Use raw workerd directly only if:**
- Self-hosting Workers runtime in production
- Embedding runtime in C++ application
- Custom tooling/testing infrastructure
- Debugging workerd-specific behavior

**Never use workerd for:**
- Running untrusted/user-submitted code
- Multi-tenant isolation (not hardened)
- Production without additional security layers

## Key Features
- **Standards-based**: Fetch API, Web Crypto, Streams, WebSocket
- **Nanoservices**: Service bindings with local call performance
- **Capability security**: Explicit bindings prevent SSRF
- **Backwards compatible**: Version = max compat date supported

## Architecture
```
Config (workerd.capnp)
├── Services (workers/endpoints)
├── Sockets (HTTP/HTTPS listeners)
└── Extensions (global capabilities)
```

## Quick Start
```bash
workerd serve config.capnp
workerd compile config.capnp myConfig -o binary
workerd test config.capnp
```

## Platform Support & Beta Status

| Platform | Status | Notes |
|----------|--------|-------|
| Linux (x64) | Stable | Primary platform |
| macOS (x64/ARM) | Stable | Full support |
| Windows | Beta | Use WSL2 for best results |
| Linux (ARM64) | Experimental | Limited testing |

workerd is in **active development**. Breaking changes possible. Pin versions in production.

## Core Concepts
- **Service**: Named endpoint (worker/network/disk/external)
- **Binding**: Capability-based resource access (KV/DO/R2/services)
- **Compatibility date**: Feature gate (always set!)
- **Modules**: ES modules (recommended) or service worker syntax

## Reading Order (Progressive Disclosure)

**Start here:**
1. This README (overview, decision tree)
2. [patterns.md](./patterns.md) - Common workflows, framework examples

**When you need details:**
3. [configuration.md](./configuration.md) - Config format, services, bindings
4. [api.md](./api.md) - Runtime APIs, TypeScript types
5. [gotchas.md](./gotchas.md) - Common errors, debugging

## Related References
- [workers](https://developers.cloudflare.com/workers/) - Workers runtime API documentation
- [miniflare](../miniflare/) - Testing tool built on workerd
- [wrangler](https://developers.cloudflare.com/workers/wrangler/) - CLI that uses workerd for local dev

# Workers AI API

Fetch the selected model's page from the [model catalog](https://developers.cloudflare.com/workers-ai/models/) for request fields, output format, dimensions, and examples. Text, embeddings, images, audio, and translation do not share one response schema.

| Task | Documentation |
|------|---------------|
| Invoke inference through a Worker binding | [Workers bindings](https://developers.cloudflare.com/workers-ai/configuration/bindings/) |
| Invoke inference over HTTP | [REST API reference](https://developers.cloudflare.com/api/resources/ai/methods/run/) |
| Stream text or use SDK abstractions | [Vercel AI SDK](https://developers.cloudflare.com/workers-ai/configuration/ai-sdk/), or the selected model's streaming example |
| Define tools and handle tool results | [Function calling](https://developers.cloudflare.com/workers-ai/features/function-calling/) |
| Request structured output | [JSON mode](https://developers.cloudflare.com/workers-ai/features/json-mode/) |

Use the response and stream format documented for the chosen integration. Do not assume native binding streams are parsed objects or apply OpenAI response parsing to every model. Check the model's batching support and limits before combining inputs in one request.

# Workers AI Configuration

Read the setup guide for the application's existing integration and installed SDK/Wrangler versions before adapting configuration.

| Task | Documentation |
|------|---------------|
| Create and develop a Worker with Workers AI | [Workers and Wrangler setup](https://developers.cloudflare.com/workers-ai/get-started/workers-wrangler/) |
| Add an AI binding to an existing Worker | [Workers bindings](https://developers.cloudflare.com/workers-ai/configuration/bindings/) |
| Generate environment and runtime types | [Workers TypeScript](https://developers.cloudflare.com/workers/languages/typescript/) |
| Call inference from outside Workers | [REST API setup and authentication](https://developers.cloudflare.com/workers-ai/get-started/rest-api/) |
| Use the Vercel AI SDK | [AI SDK integration](https://developers.cloudflare.com/workers-ai/configuration/ai-sdk/) |
| Adapt an existing OpenAI SDK client | [OpenAI compatible endpoints](https://developers.cloudflare.com/workers-ai/configuration/open-ai-compatibility/) |

Prefer the native binding for a Worker that does not need an SDK abstraction; use REST for external services. Preserve an existing SDK integration when it meets the task, and check its supported endpoints and model features before substituting providers.

Local Worker execution and local inference are different: Workers AI inference uses the Cloudflare account even during local development and consumes usage. Follow the current setup guide for development configuration; do not assume the entire Worker must run remotely.

# Workers AI Troubleshooting

Use the actual error, model identifier, integration, and installed versions to choose the relevant reference.

| Symptom or decision | Documentation and checks |
|---------------------|--------------------------|
| Missing binding or types | [Binding configuration](https://developers.cloudflare.com/workers-ai/configuration/bindings/) and [Workers TypeScript](https://developers.cloudflare.com/workers/languages/typescript/); check the environment being run |
| Development inference fails | [Workers and Wrangler setup](https://developers.cloudflare.com/workers-ai/get-started/workers-wrangler/); check account access and binding setup |
| Unknown model, invalid input, or unexpected response | Open the exact model in the [catalog](https://developers.cloudflare.com/workers-ai/models/); check its schema, context window, and feature support |
| Inference error or retry decision | [Error codes and HTTP statuses](https://developers.cloudflare.com/workers-ai/platform/errors/) |
| Throttling or concurrency planning | [Current limits](https://developers.cloudflare.com/workers-ai/platform/limits/) |
| Usage or cost estimate | [Current pricing](https://developers.cloudflare.com/workers-ai/platform/pricing/); use the selected model's billing units and expected workload |
| Old SDK examples fail | [Native binding](https://developers.cloudflare.com/workers-ai/configuration/bindings/), [AI SDK](https://developers.cloudflare.com/workers-ai/configuration/ai-sdk/), or [OpenAI compatibility](https://developers.cloudflare.com/workers-ai/configuration/open-ai-compatibility/), according to the integration |

Do not copy an error-code mapping, per-request neuron estimate, or context-window range from another model or an older example. Measure latency for the intended workload rather than promising a fixed cold-start or inference time.

# Workers AI Patterns

Use direct generation when the supplied context fits the selected model and retrieval is unnecessary. Use RAG when answers need grounding in a document corpus or relevant passages must be selected from larger data; decide from the actual model context budget rather than a fixed token threshold.

| Task | Documentation |
|------|---------------|
| Build retrieval with Workers AI, Vectorize, and document storage | [RAG tutorial](https://developers.cloudflare.com/workers-ai/guides/tutorials/build-a-retrieval-augmented-generation-ai/) |
| Stream responses or integrate tool calling in an SDK application | [AI SDK integration](https://developers.cloudflare.com/workers-ai/configuration/ai-sdk/) |
| Constrain generated JSON | [JSON mode](https://developers.cloudflare.com/workers-ai/features/json-mode/) |
| Add caching, retries, or model fallbacks | [Caching](https://developers.cloudflare.com/ai-gateway/features/caching/), [request handling](https://developers.cloudflare.com/ai-gateway/configuration/request-handling/), and [dynamic routing](https://developers.cloudflare.com/ai-gateway/features/dynamic-routing/) |

Treat tutorial models as examples; select models using the [model criteria](./README.md#choose-a-model). For RAG, embed queries and documents with compatible models and match the index dimensions to the embeddings. Budget for retrieval and embedding work as well as generation.

Before adding a fallback model, verify that it can satisfy the same schema, context, and tool requirements. For retry decisions, distinguish transient failures from invalid inputs or configuration using the [error and limit references](./gotchas.md).

# Cloudflare Workers AI

Use Workers AI for managed model inference from Workers or an external service. Fetch the relevant documentation before choosing a model or writing integration code; model availability, schemas, capabilities, limits, and prices change independently.

## Choose a model

Start with the [model catalog](https://developers.cloudflare.com/workers-ai/models/) and open the selected model's page for its exact identifier, input/output schema, context window, and supported features. Compare candidates on the user's task, language, quality requirements, latency, and [current pricing](https://developers.cloudflare.com/workers-ai/platform/pricing/). Evaluate with representative inputs rather than treating model size as a quality or cost ranking.

For tool use, streaming, or structured output, confirm support for the selected model and integration. For embeddings, check output dimensions and compatibility with the existing index; changing the model may require re-embedding stored documents, even if dimensions match.

## Route by task

- [configuration.md](./configuration.md): choose an integration, configure bindings and types, or set up development.
- [api.md](./api.md): find inference schemas, streaming, tool calling, and structured output.
- [patterns.md](./patterns.md): choose direct generation or RAG, and find integration examples.
- [gotchas.md](./gotchas.md): diagnose binding, schema, limit, pricing, and SDK issues.

If a topic is missing, use the [Workers AI documentation index](https://developers.cloudflare.com/workers-ai/llms.txt) to find its current page.

## Related products

- [Vectorize](../vectorize/): vector storage and retrieval.
- [AI Gateway](../ai-gateway/): inference analytics, caching, and request controls.
- [Workers](https://developers.cloudflare.com/workers/): runtime and application hosting.

# API Operations

## Deploy User Worker

```bash
curl -X PUT \
  "https://api.cloudflare.com/client/v4/accounts/$ACCOUNT_ID/workers/dispatch/namespaces/$NAMESPACE/scripts/$SCRIPT_NAME" \
  -H "Authorization: Bearer $API_TOKEN" \
  -F 'metadata={"main_module": "worker.mjs"};type=application/json' \
  -F 'worker.mjs=@worker.mjs;type=application/javascript+module'
```

### TypeScript SDK
```typescript
import Cloudflare from "cloudflare";

const client = new Cloudflare({ apiToken: process.env.API_TOKEN });

const scriptFile = new File([scriptContent], `${scriptName}.mjs`, {
  type: "application/javascript+module",
});

await client.workersForPlatforms.dispatch.namespaces.scripts.update(
  namespace, scriptName,
  {
    account_id: accountId,
    metadata: { main_module: `${scriptName}.mjs` },
    files: [scriptFile],
  }
);
```

## TypeScript Types

```typescript
import type { DispatchNamespace } from '@cloudflare/workers-types';

interface DispatchNamespace {
  get(name: string, options?: Record<string, unknown>, dispatchOptions?: DynamicDispatchOptions): Fetcher;
}

interface DynamicDispatchOptions {
  limits?: DynamicDispatchLimits;
  outbound?: Record<string, unknown>;
}

interface DynamicDispatchLimits {
  cpuMs?: number;        // Max CPU milliseconds
  subRequests?: number;  // Max fetch() calls
}

// Usage
const userWorker = env.DISPATCHER.get('customer-123', {}, {
  limits: { cpuMs: 50, subRequests: 20 },
  outbound: { customerId: '123', url: request.url }
});
```

## Deploy with Bindings
```bash
curl -X PUT ".../scripts/$SCRIPT_NAME" \
  -F 'metadata={
    "main_module": "worker.mjs",
    "bindings": [
      {"type": "kv_namespace", "name": "MY_KV", "namespace_id": "'$KV_ID'"}
    ],
    "tags": ["customer-123", "production"],
    "compatibility_date": "2026-01-01"  // Use current date for new projects
  };type=application/json' \
  -F 'worker.mjs=@worker.mjs;type=application/javascript+module'
```

## List/Delete Workers

```bash
# List
curl "https://api.cloudflare.com/client/v4/accounts/$ACCOUNT_ID/workers/dispatch/namespaces/$NAMESPACE/scripts" \
  -H "Authorization: Bearer $API_TOKEN"

# Delete by name
curl -X DELETE ".../scripts/$SCRIPT_NAME" -H "Authorization: Bearer $API_TOKEN"

# Delete by tag
curl -X DELETE ".../scripts?tags=customer-123%3Ayes" -H "Authorization: Bearer $API_TOKEN"
```

**Pagination:** SDK supports async iteration. Manual: add `?per_page=100&page=1` query params.

## Static Assets

**3-step process:** Create session → Upload files → Deploy Worker

### 1. Create Upload Session
```bash
curl -X POST ".../scripts/$SCRIPT_NAME/assets-upload-session" \
  -H "Authorization: Bearer $API_TOKEN" \
  -d '{
    "manifest": {
      "/index.html": {"hash": "08f1dfda4574284ab3c21666d1ee8c7d4", "size": 1234}
    }
  }'
# Returns: jwt, buckets
```

**Hash:** SHA-256 truncated to first 16 bytes (32 hex characters)

### 2. Upload Files
```bash
curl -X POST ".../workers/assets/upload?base64=true" \
  -H "Authorization: Bearer $UPLOAD_JWT" \
  -F '08f1dfda4574284ab3c21666d1ee8c7d4=<BASE64_CONTENT>'
# Returns: completion jwt
```

**Multiple buckets:** Upload to all returned bucket URLs (typically 2 for redundancy) using same JWT and hash.

### 3. Deploy with Assets
```bash
curl -X PUT ".../scripts/$SCRIPT_NAME" \
  -F 'metadata={
    "main_module": "index.js",
    "assets": {"jwt": "<COMPLETION_TOKEN>"},
    "bindings": [{"type": "assets", "name": "ASSETS"}]
  };type=application/json' \
  -F 'index.js=export default {...};type=application/javascript+module'
```

**Asset Isolation:** Assets shared across namespace by default. For customer isolation, salt hash: `sha256(customerId + fileContents).slice(0, 32)`

## Dispatch Workers

### Subdomain Routing
```typescript
export default {
  async fetch(request: Request, env: Env): Promise<Response> {
    const userWorkerName = new URL(request.url).hostname.split(".")[0];
    const userWorker = env.DISPATCHER.get(userWorkerName);
    return await userWorker.fetch(request);
  },
};
```

### Path Routing
```typescript
const pathParts = new URL(request.url).pathname.split("/").filter(Boolean);
const userWorker = env.DISPATCHER.get(pathParts[0]);
return await userWorker.fetch(request);
```

### KV Routing
```typescript
const hostname = new URL(request.url).hostname;
const userWorkerName = await env.ROUTING_KV.get(hostname);
const userWorker = env.DISPATCHER.get(userWorkerName);
return await userWorker.fetch(request);
```

## Outbound Workers

Control external fetch from user Workers:

### Configure
```typescript
const userWorker = env.DISPATCHER.get(
  workerName, {},
  { outbound: { customer_context: { customer_name: workerName, url: request.url } } }
);
```

### Implement
```typescript
export default {
  async fetch(request: Request, env: Env): Promise<Response> {
    const customerName = env.customer_name;
    const url = new URL(request.url);
    
    // Block domains
    if (["malicious.com"].some(d => url.hostname.includes(d))) {
      return new Response("Blocked", { status: 403 });
    }
    
    // Inject auth
    if (url.hostname === "api.example.com") {
      const headers = new Headers(request.headers);
      headers.set("Authorization", `Bearer ${generateJWT(customerName)}`);
      return fetch(new Request(request, { headers }));
    }
    
    return fetch(request);
  },
};
```

**Note:** Doesn't intercept DO/mTLS fetch.

See [README.md](./README.md), [configuration.md](./configuration.md), [patterns.md](./patterns.md), [gotchas.md](./gotchas.md)

# Configuration

## Dispatch Namespace Binding

### wrangler.jsonc
```jsonc
{
  "$schema": "./node_modules/wrangler/config-schema.json",
  "dispatch_namespaces": [{
    "binding": "DISPATCHER",
    "namespace": "production"
  }]
}
```

## Worker Isolation Mode

Workers in a namespace run in **untrusted mode** by default for security:
- No access to `request.cf` object
- Isolated cache per Worker (no shared cache)
- `caches.default` disabled

### Enable Trusted Mode

For internal platforms where you control all code:

```bash
curl -X PUT \
  "https://api.cloudflare.com/client/v4/accounts/$ACCOUNT_ID/workers/dispatch/namespaces/$NAMESPACE" \
  -H "Authorization: Bearer $API_TOKEN" \
  -d '{"name": "'$NAMESPACE'", "trusted_workers": true}'
```

**Caveats:**
- Workers share cache within namespace (use cache key prefixes: `customer-${id}:${key}`)
- `request.cf` object accessible
- Redeploy existing Workers after enabling trusted mode

**When to use:** Internal platforms, A/B testing platforms, need geolocation data


### With Outbound Worker
```jsonc
{
  "dispatch_namespaces": [{
    "binding": "DISPATCHER",
    "namespace": "production",
    "outbound": {
      "service": "outbound-worker",
      "parameters": ["customer_context"]
    }
  }]
}
```

## Wrangler Commands

```bash
wrangler dispatch-namespace list
wrangler dispatch-namespace get production
wrangler dispatch-namespace create production
wrangler dispatch-namespace delete staging
wrangler dispatch-namespace rename old new
```

## Custom Limits

Set CPU time and subrequest limits per invocation:

```typescript
const userWorker = env.DISPATCHER.get(
  workerName,
  {},
  {
    limits: { 
      cpuMs: 10,        // Max CPU ms
      subRequests: 5    // Max fetch() calls
    }
  }
);
```

Handle limit violations:
```typescript
try {
  return await userWorker.fetch(request);
} catch (e) {
  if (e.message.includes("CPU time limit")) {
    return new Response("CPU limit exceeded", { status: 429 });
  }
  throw e;
}
```

## Static Assets

Deploy HTML/CSS/images with Workers. See [api.md](./api.md#static-assets) for upload process.

### Wrangler
```jsonc
{
  "name": "customer-site",
  "main": "./src/index.js",
  "assets": {
    "directory": "./public",
    "binding": "ASSETS"
  }
}
```

```bash
npx wrangler deploy --name customer-site --dispatch-namespace production
```

### Dashboard Deployment

Alternative to CLI:

1. Upload Worker file in dashboard
2. Add `--dispatch-namespace` flag: `wrangler deploy --dispatch-namespace production`
3. Or configure in wrangler.jsonc under `dispatch_namespaces`

See [api.md](./api.md) for programmatic deployment via REST API or SDK.

## Tags

Organize/search Workers (max 8/script):

```bash
# Set tags
curl -X PUT ".../tags" -d '["customer-123", "pro", "production"]'

# Filter by tag
curl ".../scripts?tags=production%3Ayes"

# Delete by tag
curl -X DELETE ".../scripts?tags=customer-123%3Ayes"
```

Common patterns: `customer-123`, `free|pro|enterprise`, `production|staging`

## Bindings

**Supported binding types:** 29 total including KV, D1, R2, Durable Objects, Analytics Engine, Service, Assets, Queue, Vectorize, Hyperdrive, Workflow, AI, Browser, and more.

Add via API metadata (see [api.md](./api.md#deploy-with-bindings)):
```json
{
  "bindings": [
    {"type": "kv_namespace", "name": "USER_KV", "namespace_id": "..."},
    {"type": "r2_bucket", "name": "STORAGE", "bucket_name": "..."},
    {"type": "d1", "name": "DB", "id": "..."}
  ]
}
```

Preserve existing bindings:
```json
{
  "bindings": [{"type": "r2_bucket", "name": "STORAGE", "bucket_name": "new"}],
  "keep_bindings": ["kv_namespace", "d1"]  // Preserves existing bindings of these types
}
```

For complete binding type reference, see [bindings](../bindings/) documentation

See [README.md](./README.md), [api.md](./api.md), [patterns.md](./patterns.md), [gotchas.md](./gotchas.md)

# Gotchas & Limits

## Common Errors

### "Worker not found"

**Cause:** Attempting to get Worker that doesn't exist in namespace  
**Solution:** Catch error and return 404:

```typescript
try {
  const userWorker = env.DISPATCHER.get(workerName);
  return userWorker.fetch(request);
} catch (e) {
  if (e.message.startsWith("Worker not found")) {
    return new Response("Worker not found", { status: 404 });
  }
  throw e;  // Re-throw unexpected errors
}
```

### "CPU time limit exceeded"

**Cause:** User Worker exceeded configured CPU time limit  
**Solution:** Track violations in Analytics Engine and return 429 response; consider adjusting limits per customer tier

### "Hostname Routing Issues"

**Cause:** DNS proxy settings causing routing problems  
**Solution:** Use `*/*` wildcard route which works regardless of proxy settings for orange-to-orange routing

### "Bindings Lost on Update"

**Cause:** Not using `keep_bindings` flag when updating Worker  
**Solution:** Use `keep_bindings: true` in API requests to preserve existing bindings during updates

### "Tag Filtering Not Working"

**Cause:** Special characters not URL encoded in tag filters  
**Solution:** URL encode tags (e.g., `tags=production%3Ayes`) and avoid special chars like `,` and `&`

### "Deploy Failures with ES Modules"

**Cause:** Incorrect upload format for ES modules  
**Solution:** Use multipart form upload, specify `main_module` in metadata, and set file type to `application/javascript+module`

### "Static Asset Upload Failed"

**Cause:** Invalid hash format, expired token, or incorrect encoding  
**Solution:** Hash must be first 16 bytes (32 hex chars) of SHA-256, upload within 1 hour of session creation, deploy within 1 hour of upload completion, and Base64 encode file contents

### "Outbound Worker Not Intercepting Calls"

**Cause:** Outbound Workers don't intercept Durable Object or mTLS binding fetch  
**Solution:** Plan egress control accordingly; not all fetch calls are intercepted

### "TCP Socket Connection Failed"

**Cause:** Outbound Worker enabled blocks `connect()` API for TCP sockets  
**Solution:** Outbound Workers only intercept `fetch()` calls; TCP socket connections unavailable when outbound configured. Remove outbound if TCP needed, or use proxy pattern.

### "API Rate Limit Exceeded"

**Cause:** Exceeded Cloudflare API rate limits (1200 requests per 5 minutes per account, 200 requests per second per IP)  
**Solution:** Implement exponential backoff:

```typescript
async function deployWithBackoff(deploy: () => Promise<void>, maxRetries = 3) {
  for (let i = 0; i < maxRetries; i++) {
    try {
      return await deploy();
    } catch (e) {
      if (e.status === 429 && i < maxRetries - 1) {
        await new Promise(r => setTimeout(r, Math.pow(2, i) * 1000));
        continue;
      }
      throw e;
    }
  }
}
```

### "Gradual Deployment Not Supported"

**Cause:** Attempted to use gradual deployments with user Workers  
**Solution:** Gradual deployments not supported for Workers in dispatch namespaces. Use all-at-once deployment with staged rollout via dispatch worker logic (feature flags, percentage-based routing).

### "Asset Session Expired"

**Cause:** Upload JWT expired (1 hour validity) or completion token expired (1 hour after upload)  
**Solution:** Complete asset upload within 1 hour of session creation, and deploy Worker within 1 hour of upload completion. For large uploads, batch files or increase upload parallelism.

## Platform Limits

| Limit | Value | Notes |
|-------|-------|-------|
| Workers per namespace | Unlimited | Unlike regular Workers (500 per account) |
| Namespaces per account | Unlimited | Best practice: 1 production + 1 staging |
| Max tags per Worker | 8 | For filtering and organization |
| Worker mode | Untrusted (default) | No `request.cf` access unless trusted mode |
| Cache isolation | Per-Worker (untrusted) | Shared in trusted mode with key prefixes |
| Durable Object namespaces | Unlimited | No per-account limit for WfP |
| Gradual Deployments | Not supported | All-at-once only |
| `caches.default` | Disabled (untrusted) | Use Cache API with custom keys |

## Asset Upload Limits

| Limit | Value | Notes |
|-------|-------|-------|
| Upload session JWT validity | 1 hour | Must complete upload within this time |
| Completion token validity | 1 hour | Must deploy within this time after upload |
| Asset hash format | First 16 bytes SHA-256 | 32 hex characters |
| Base64 encoding | Required | For binary files |

## API Rate Limits

| Limit Type | Value | Scope |
|------------|-------|-------|
| Client API | 1200 requests / 5 min | Per account |
| Client API | 200 requests / sec | Per IP address |
| GraphQL | Varies by query cost | Query complexity |

See [Cloudflare API Rate Limits](https://developers.cloudflare.com/fundamentals/api/reference/limits/) for details.

## Operational Limits

| Operation | Limit | Notes |
|-----------|-------|-------|
| CPU time (custom limits) | Up to Workers plan limit | Set per-invocation in dispatch worker |
| Subrequests (custom limits) | Up to Workers plan limit | Set per-invocation in dispatch worker |
| Outbound Worker subrequests | Not intercepted for DO/mTLS | Only regular fetch() calls |
| TCP sockets with outbound | Disabled | `connect()` API unavailable |

See [README.md](./README.md), [configuration.md](./configuration.md), [api.md](./api.md), [patterns.md](./patterns.md)

# Multi-Tenant Patterns

## Billing by Plan

```typescript
interface Env {
  DISPATCHER: DispatchNamespace;
  CUSTOMERS_KV: KVNamespace;
}

export default {
  async fetch(request: Request, env: Env): Promise<Response> {
    const userWorkerName = new URL(request.url).hostname.split(".")[0];
    const customerPlan = await env.CUSTOMERS_KV.get(userWorkerName);
    
    const plans = {
      enterprise: { cpuMs: 50, subRequests: 50 },
      pro: { cpuMs: 20, subRequests: 20 },
      free: { cpuMs: 10, subRequests: 5 },
    };
    const limits = plans[customerPlan as keyof typeof plans] || plans.free;
    
    const userWorker = env.DISPATCHER.get(userWorkerName, {}, { limits });
    return await userWorker.fetch(request);
  },
};
```

## Resource Isolation

**Complete isolation:** Create unique resources per customer
- KV namespace per customer
- D1 database per customer
- R2 bucket per customer

```typescript
const bindings = [{
  type: "kv_namespace",
  name: "USER_KV",
  namespace_id: `customer-${customerId}-kv`
}];
```

## Hostname Routing

### Wildcard Route (Recommended)
Configure `*/*` route on SaaS domain → dispatch Worker

**Benefits:**
- Supports subdomains + custom vanity domains
- No per-route limits (regular Workers limited to 100 routes)
- Programmatic control
- Works with any DNS proxy settings

**Setup:**
1. Cloudflare for SaaS custom hostnames
2. Fallback origin (dummy `A 192.0.2.0` if Worker is origin)
3. DNS CNAME to SaaS domain
4. `*/*` route → dispatch Worker
5. Routing logic in dispatch Worker

```typescript
export default {
  async fetch(request: Request, env: Env): Promise<Response> {
    const hostname = new URL(request.url).hostname;
    const hostnameData = await env.ROUTING_KV.get(`hostname:${hostname}`, { type: "json" });
    
    if (!hostnameData?.workerName) {
      return new Response("Hostname not configured", { status: 404 });
    }
    
    const userWorker = env.DISPATCHER.get(hostnameData.workerName);
    return await userWorker.fetch(request);
  },
};
```

### Subdomain-Only
1. Wildcard DNS: `*.saas.com` → origin
2. Route: `*.saas.com/*` → dispatch Worker
3. Extract subdomain for routing

### Orange-to-Orange (O2O) Behavior

When customers use Cloudflare and CNAME to your Workers domain:

| Scenario | Behavior | Route Pattern |
|----------|----------|---------------|
| Customer not on Cloudflare | Standard routing | `*/*` or `*.domain.com/*` |
| Customer on Cloudflare (proxied CNAME) | Invokes Worker at edge | `*/*` required |
| Customer on Cloudflare (DNS-only CNAME) | Standard routing | Any route works |

**Recommendation:** Always use `*/*` wildcard for consistent O2O behavior.

### Custom Metadata Routing

For Cloudflare for SaaS: Store worker name in custom hostname `custom_metadata`, retrieve in dispatch worker to route requests. Requires custom hostnames as subdomains of your domain.

## Observability

### Logpush
- Enable on dispatch Worker → captures all user Worker logs
- Filter by `Outcome` or `Script Name`

### Tail Workers
- Real-time logs with custom formatting
- Receives HTTP status, `console.log()`, exceptions, diagnostics

### Analytics Engine
```typescript
// Track violations
env.ANALYTICS.writeDataPoint({
  indexes: [customerName],
  blobs: ["cpu_limit_exceeded"],
});
```

### GraphQL
```graphql
query {
  viewer {
    accounts(filter: {accountTag: $accountId}) {
      workersInvocationsAdaptive(filter: {dispatchNamespaceName: "production"}) {
        sum { requests errors cpuTime }
      }
    }
  }
}
```

## Use Case Implementations

### AI Code Execution
```typescript
async function deployGeneratedCode(name: string, code: string) {
  const file = new File([code], `${name}.mjs`, { type: "application/javascript+module" });
  await client.workersForPlatforms.dispatch.namespaces.scripts.update("production", name, {
    account_id: accountId,
    metadata: { main_module: `${name}.mjs`, tags: [name, "ai-generated"] },
    files: [file],
  });
}

// Short limits for untrusted code
const userWorker = env.DISPATCHER.get(sessionId, {}, { limits: { cpuMs: 5, subRequests: 3 } });
```

**VibeSDK:** For AI-powered code generation + deployment platforms, see [VibeSDK](https://github.com/cloudflare/vibesdk) - handles AI generation, sandbox execution, live preview, and deployment.

Reference: [AI Vibe Coding Platform Architecture](https://developers.cloudflare.com/reference-architecture/diagrams/ai/ai-vibe-coding-platform/)

### Edge Functions Platform
```typescript
// Route: /customer-id/function-name
const [customerId, functionName] = new URL(request.url).pathname.split("/").filter(Boolean);
const workerName = `${customerId}-${functionName}`;
const userWorker = env.DISPATCHER.get(workerName);
```

### Website Builder
- Deploy static assets + Worker code
- See [api.md](./api.md#static-assets) for full implementation
- Salt hashes for asset isolation

## Best Practices

### Architecture
- One namespace per environment (production, staging)
- Platform logic in dispatch Worker (auth, rate limiting, validation)
- Isolation automatic (no shared cache, untrusted mode)

### Routing
- Use `*/*` wildcard routes
- Store mappings in KV
- Handle missing Workers gracefully

### Limits & Security
- Set custom limits by plan
- Track violations with Analytics Engine
- Use outbound Workers for egress control
- Sanitize responses

### Tags
- Tag all Workers: customer ID, plan, environment
- Enable bulk operations
- Filter efficiently

See [README.md](./README.md), [configuration.md](./configuration.md), [api.md](./api.md), [gotchas.md](./gotchas.md)

# Cloudflare Workers for Platforms

Multi-tenant platform with isolated customer code execution at scale.

## Use Cases

- Multi-tenant SaaS running customer code
- AI-generated code execution in secure sandboxes
- Programmable platforms with isolated compute
- Edge functions/serverless platforms
- Website builders with static + dynamic content
- Unlimited app deployment at scale

**NOT for general Workers** - only for Workers for Platforms architecture.

## Quick Start

**One-click deploy:** [Platform Starter Kit](https://github.com/cloudflare/workers-for-platforms-example) deploys complete WfP setup with dispatch namespace, dispatch worker, and user worker example.

[![Deploy to Cloudflare](https://deploy.workers.cloudflare.com/button)](https://deploy.workers.cloudflare.com/?url=https://github.com/cloudflare/workers-for-platforms-example)

**Manual setup:** See [configuration.md](./configuration.md) for namespace creation and dispatch worker configuration.

## Key Features

- Unlimited Workers per namespace (no script limits)
- Automatic tenant isolation
- Custom CPU/subrequest limits per customer
- Hostname routing (subdomains/vanity domains)
- Egress/ingress control
- Static assets support
- Tags for bulk operations

## Architecture

**4 Components:**
1. **Dispatch Namespace** - Container for unlimited customer Workers, automatic isolation (untrusted mode by default - no request.cf access, no shared cache)
2. **Dynamic Dispatch Worker** - Entry point, routes requests, enforces platform logic (auth, limits, validation)
3. **User Workers** - Customer code in isolated sandboxes, API-deployed, optional bindings (KV/D1/R2/DO)
4. **Outbound Worker** (optional) - Intercepts external fetch, controls egress, logs subrequests (blocks TCP socket connect() API)

**Request Flow:**
```
Request → Dispatch Worker → Determines user Worker → env.DISPATCHER.get("customer") 
→ User Worker executes (Outbound Worker for external fetch) → Response → Dispatch Worker → Client
```

## Decision Trees

### When to Use Workers for Platforms
```
Need to run code?
├─ Your code only → Regular Workers
├─ Customer/AI code → Workers for Platforms
└─ Untrusted code in sandbox → Workers for Platforms OR Sandbox API
```

### Routing Strategy Selection
```
Hostname routing needed?
├─ Subdomains only (*.saas.com) → `*.saas.com/*` route + subdomain extraction
├─ Custom domains → `*/*` wildcard + Cloudflare for SaaS + KV/metadata routing
└─ Path-based (/customer/app) → Any route + path parsing
```

### Isolation Mode Selection
```
Worker mode?
├─ Running customer code → Untrusted (default)
├─ Need request.cf geolocation → Trusted mode
├─ Internal platform, controlled code → Trusted mode with cache key prefixes
└─ Maximum isolation → Untrusted + unique resources per customer
```

## In This Reference

| File | Purpose | When to Read |
|------|---------|--------------|
| [configuration.md](./configuration.md) | Namespace setup, dispatch worker config | First-time setup, changing limits |
| [api.md](./api.md) | User worker API, dispatch API, outbound worker | Deploying workers, SDK integration |
| [patterns.md](./patterns.md) | Multi-tenancy, routing, egress control | Planning architecture, scaling |
| [gotchas.md](./gotchas.md) | Limits, isolation issues, best practices | Debugging, production prep |

## See Also
- [workers](https://developers.cloudflare.com/workers/) - Core Workers runtime documentation
- [durable-objects](https://developers.cloudflare.com/durable-objects/) - Stateful multi-tenant patterns
- [sandbox](https://developers.cloudflare.com/sandbox/) - Alternative for untrusted code execution
- [Reference Architecture: Programmable Platforms](https://developers.cloudflare.com/reference-architecture/diagrams/serverless/programmable-platforms/)
- [Reference Architecture: AI Vibe Coding Platform](https://developers.cloudflare.com/reference-architecture/diagrams/ai/ai-vibe-coding-platform/)

# Workers Playground API

## Handler

```javascript
export default {
  async fetch(request, env, ctx) {
    // request: Request, env: {} (empty in playground), ctx: ExecutionContext
    return new Response('Hello');
  }
};
```

## Request

```javascript
const method = request.method;       // "GET", "POST"
const url = new URL(request.url);    // Parse URL
const headers = request.headers;     // Headers object
const body = await request.json();   // Read body (consumes stream)
const clone = request.clone();       // Clone before reading body

// Query params
url.searchParams.get('page');        // Single value
url.searchParams.getAll('tag');      // Array

// Cloudflare metadata
request.cf.country;                  // "US"
request.cf.colo;                     // "SFO"
```

## Response

```javascript
// Text
return new Response('Hello', { status: 200 });

// JSON
return Response.json({ data }, { status: 200, headers: {...} });

// Redirect
return Response.redirect('/new-path', 301);

// Modify existing
const modified = new Response(response.body, response);
modified.headers.set('X-Custom', 'value');
```

## ExecutionContext

```javascript
// Background work (after response sent)
ctx.waitUntil(fetch('https://logs.example.com', { method: 'POST', body: '...' }));
return new Response('OK'); // Returns immediately
```

## Fetch

```javascript
const response = await fetch('https://api.example.com');
const data = await response.json();

// With options
await fetch(url, {
  method: 'POST',
  headers: { 'Content-Type': 'application/json' },
  body: JSON.stringify({ name: 'Alice' })
});
```

## Cache

```javascript
const cache = caches.default;

// Check cache
let response = await cache.match(request);
if (!response) {
  response = await fetch(origin);
  await cache.put(request, response.clone()); // Clone before put!
}
return response;
```

## Crypto

```javascript
crypto.randomUUID();                 // UUID v4
crypto.getRandomValues(new Uint8Array(16));

// SHA-256 hash
const hash = await crypto.subtle.digest('SHA-256', new TextEncoder().encode(data));
```

## Limits (Playground = Free Plan)

| Resource | Limit |
|----------|-------|
| CPU time | 10ms (Free plan; Paid: 30s default, 5min max) |
| Subrequests | 50 |
| Memory | 128 MB |

# Configuration

## Getting Started

Navigate to [workers.cloudflare.com/playground](https://workers.cloudflare.com/playground)

- **No account required** for testing
- **No CLI or local setup** needed
- Code executes in real Cloudflare Workers runtime
- Share code via URL (never expires)

## Playground Constraints

⚠️ **Important Limitations**

| Constraint | Playground | Production Workers |
|------------|------------|-------------------|
| **Module Format** | ES modules only | ES modules or Service Worker |
| **TypeScript** | Not supported (JS only) | Supported via build step |
| **Bindings** | Not available | KV, D1, R2, Durable Objects, etc. |
| **wrangler.toml** | Not used | Required for config |
| **Environment Variables** | Not available | Full support |
| **Secrets** | Not available | Full support |
| **Custom Domains** | Not available | Full support |

**Playground is for rapid prototyping only.** For production apps, use `wrangler` CLI.

## Code Editor

### Syntax Requirements

Must export default object with `fetch` handler:

```javascript
export default {
  async fetch(request, env, ctx) {
    return new Response('Hello World');
  }
};
```

**Key Points:**
- Must use ES modules (`export default`)
- `fetch` method receives `(request, env, ctx)`
- Must return `Response` object
- TypeScript not supported (use plain JavaScript)

### Multi-Module Code

Import from external URLs or inline modules:

```javascript
// Import from CDN
import { Hono } from 'https://esm.sh/hono@3';

// Or paste library code and import relatively
// (See patterns.md for multi-module examples)

export default {
  async fetch(request) {
    const app = new Hono();
    app.get('/', (c) => c.text('Hello'));
    return app.fetch(request);
  }
};
```

## Preview Panel

### Browser Tab

Default interactive preview with address bar:
- Enter custom URL paths
- Automatic reload on code changes
- DevTools available (right-click → Inspect)

### HTTP Test Panel

Switch to **HTTP** tab for raw HTTP testing:
- Change HTTP method (GET, POST, PUT, DELETE, PATCH, etc.)
- Add/edit request headers
- Modify request body (JSON, form data, text)
- View response headers and body
- Test different content types

Example HTTP test:
```
Method: POST
URL: /api/users
Headers:
  Content-Type: application/json
  Authorization: Bearer token123
Body:
{
  "name": "Alice",
  "email": "alice@example.com"
}
```

## Sharing Code

**Copy Link** button generates shareable URL:
- Code embedded in URL fragment
- Links never expire
- No account required
- Can be bookmarked for later

Example: `https://workers.cloudflare.com/playground#abc123...`

## Deploying from Playground

Click **Deploy** button to move code to production:

1. **Log in** to Cloudflare account (creates free account if needed)
2. **Review** Worker name and code
3. **Deploy** to global network (takes ~30 seconds)
4. **Get URL**: Deployed to `<name>.workers.dev` subdomain
5. **Manage** from dashboard: add bindings, custom domains, analytics

**After deploy:**
- Code runs on Cloudflare's global network (300+ cities)
- Can add KV, D1, R2, Durable Objects bindings
- Configure custom domains and routes
- View analytics and logs
- Set environment variables and secrets

**Note:** Deployed Workers are production-ready but start on Free plan (100k requests/day).

## Browser Compatibility

| Browser | Status | Notes |
|---------|--------|-------|
| Chrome/Edge | ✅ Full support | Recommended |
| Firefox | ✅ Full support | Works well |
| Safari | ⚠️ Broken | Preview fails with "PreviewRequestFailed" |

**Safari users:** Use Chrome, Firefox, or Edge for Workers Playground.

## DevTools Integration

1. **Open preview** in browser tab
2. **Right-click** → Inspect Element
3. **Console tab** shows Worker logs:
   - `console.log()` output
   - Uncaught errors
   - Network requests (subrequests)

**Note:** DevTools show client-side console, not Worker execution logs. For production logging, use Logpush or Tail Workers.

## Limits in Playground

Same as production Free plan:

| Resource | Limit | Notes |
|----------|-------|-------|
| CPU time | 10ms | Per request |
| Memory | 128 MB | Per request |
| Script size | 1 MB | After compression |
| Subrequests | 50 | Outbound fetch calls |
| Request size | 100 MB | Incoming |
| Response size | Unlimited | Outgoing (streamed) |

**Exceeding CPU time** throws error immediately. Optimize hot paths or upgrade to Paid plan (30s default, 5min max CPU).

# Workers Playground Gotchas

## Platform Limitations

| Limitation | Impact | Workaround |
|------------|--------|------------|
| Safari broken | Preview fails | Use Chrome/Firefox/Edge |
| TypeScript unsupported | TS syntax errors | Write plain JS or use JSDoc |
| No bindings | `env` always `{}` | Mock data or use external APIs |
| No env vars | Can't access secrets | Hardcode for testing |

## Common Runtime Errors

### "Response body already read"

```javascript
// ❌ Body consumed twice
const body = await request.text();
await fetch(url, { body: request.body }); // Error!

// ✅ Clone first
const clone = request.clone();
const body = await request.text();
await fetch(url, { body: clone.body });
```

### "Worker exceeded CPU time"

**Limit:** 10ms (free), 30s default / 5min max (paid)

```javascript
// ✅ Move slow work to background
ctx.waitUntil(fetch('https://analytics.example.com', {...}));
return new Response('OK'); // Return immediately
```

### "Too many subrequests"

**Limit:** 50 (free), 1000 (paid)

```javascript
// ❌ 100 individual fetches
// ✅ Batch into single API call
await fetch('https://api.example.com/batch', {
  body: JSON.stringify({ ids: [...] })
});
```

## Best Practices

```javascript
// Clone before caching
await cache.put(request, response.clone());
return response;

// Validate input early
if (request.method !== 'POST') return new Response('', { status: 405 });

// Handle errors
try { ... } catch (e) {
  return Response.json({ error: e.message }, { status: 500 });
}
```

## Limits

| Resource | Free | Paid |
|----------|------|------|
| CPU time | 10ms | 30s (default), 5min (max) |
| Memory | 128 MB | 128 MB |
| Subrequests | 50 | 10,000 |

## Browser Support

| Browser | Status |
|---------|--------|
| Chrome | ✅ Recommended |
| Firefox | ✅ Works |
| Edge | ✅ Works |
| Safari | ❌ Broken |

## Debugging

```javascript
console.log('URL:', request.url); // View in browser DevTools Console
```

**Note:** `console.log` works in playground. For production, use Logpush or Tail Workers.

# Workers Playground Patterns

## JSON API

```javascript
export default {
  async fetch(request) {
    const url = new URL(request.url);
    if (url.pathname === '/api/hello') return Response.json({ message: 'Hello' });
    if (url.pathname === '/api/echo' && request.method === 'POST') {
      return Response.json({ received: await request.json() });
    }
    return Response.json({ error: 'Not found' }, { status: 404 });
  }
};
```

## Router Pattern

```javascript
const routes = {
  '/': () => new Response('Home'),
  '/api/users': () => Response.json([{ id: 1, name: 'Alice' }])
};

export default {
  async fetch(request) {
    const handler = routes[new URL(request.url).pathname];
    return handler ? handler() : new Response('Not Found', { status: 404 });
  }
};
```

## Proxy Pattern

```javascript
export default {
  async fetch(request) {
    const url = new URL(request.url);
    url.hostname = 'api.example.com';
    return fetch(url.toString(), {
      method: request.method, headers: request.headers, body: request.body
    });
  }
};
```

## CORS Handling

```javascript
export default {
  async fetch(request) {
    if (request.method === 'OPTIONS') {
      return new Response(null, {
        headers: {
          'Access-Control-Allow-Origin': '*',
          'Access-Control-Allow-Methods': 'GET, POST, PUT, DELETE',
          'Access-Control-Allow-Headers': 'Content-Type, Authorization'
        }
      });
    }
    const response = await fetch('https://api.example.com', request);
    const modified = new Response(response.body, response);
    modified.headers.set('Access-Control-Allow-Origin', '*');
    return modified;
  }
};
```

## Caching

```javascript
export default {
  async fetch(request) {
    if (request.method !== 'GET') return fetch(request);
    const cache = caches.default;
    let response = await cache.match(request);
    if (!response) {
      response = await fetch('https://api.example.com');
      if (response.status === 200) await cache.put(request, response.clone());
    }
    return response;
  }
};
```

## Hono Framework

```javascript
import { Hono } from 'https://esm.sh/hono@3';
const app = new Hono();
app.get('/', (c) => c.text('Hello'));
app.get('/api/users/:id', (c) => c.json({ id: c.req.param('id') }));
app.notFound((c) => c.json({ error: 'Not found' }, 404));
export default app;
```

## Authentication

```javascript
export default {
  async fetch(request) {
    const auth = request.headers.get('Authorization');
    if (!auth?.startsWith('Bearer ')) {
      return Response.json({ error: 'Unauthorized' }, { status: 401 });
    }
    const token = auth.substring(7);
    if (token !== 'secret-token') {
      return Response.json({ error: 'Invalid token' }, { status: 403 });
    }
    return Response.json({ message: 'Authenticated' });
  }
};
```

## Error Handling

```javascript
export default {
  async fetch(request) {
    try {
      const response = await fetch('https://api.example.com');
      if (!response.ok) throw new Error(`API returned ${response.status}`);
      return response;
    } catch (error) {
      return Response.json({ error: error.message }, { status: 500 });
    }
  }
};
```

**Note:** In-memory state (Maps, variables) resets on Worker cold start. Use Durable Objects or KV for persistence.

# Cloudflare Workers Playground Skill Reference

## Overview

Cloudflare Workers Playground is a browser-based sandbox for instantly experimenting with, testing, and deploying Cloudflare Workers without authentication or setup. This skill provides patterns, APIs, and best practices specifically for Workers Playground development.

**URL:** [workers.cloudflare.com/playground](https://workers.cloudflare.com/playground)

## ⚠️ Playground Constraints

**Playground is NOT production-equivalent:**
- ✅ Real Workers runtime, instant testing, shareable URLs
- ❌ No TypeScript (JavaScript only)
- ❌ No bindings (KV, D1, R2, Durable Objects)
- ❌ No environment variables or secrets
- ❌ ES modules only (no Service Worker format)
- ⚠️ Safari broken (use Chrome/Firefox)

**For production:** Use `wrangler` CLI. Playground is for rapid prototyping.

## Quick Start

Minimal Worker:

```javascript
export default {
  async fetch(request, env, ctx) {
    return new Response('Hello World');
  }
};
```

JSON API:

```javascript
export default {
  async fetch(request, env, ctx) {
    const data = { message: 'Hello', timestamp: Date.now() };
    return Response.json(data);
  }
};
```

Proxy with modification:

```javascript
export default {
  async fetch(request, env, ctx) {
    const response = await fetch('https://example.com');
    const modified = new Response(response.body, response);
    modified.headers.set('X-Custom-Header', 'added-by-worker');
    return modified;
  }
};
```

Import from CDN:

```javascript
import { Hono } from 'https://esm.sh/hono@3';

export default {
  async fetch(request) {
    const app = new Hono();
    app.get('/', (c) => c.text('Hello Hono!'));
    return app.fetch(request);
  }
};
```

## Reading Order

1. **[configuration.md](configuration.md)** - Start here: playground setup, constraints, deployment
2. **[api.md](api.md)** - Core APIs: Request, Response, ExecutionContext, fetch, Cache
3. **[patterns.md](patterns.md)** - Common use cases: routing, proxying, A/B testing, multi-module code
4. **[gotchas.md](gotchas.md)** - Troubleshooting: errors, browser issues, limits, best practices

## In This Reference

- **[configuration.md](configuration.md)** - Setup, deployment, configuration
- **[api.md](api.md)** - API endpoints, methods, interfaces
- **[patterns.md](patterns.md)** - Common patterns, use cases, examples
- **[gotchas.md](gotchas.md)** - Troubleshooting, best practices, limitations

## Key Features

**No Setup Required:**
- Open URL and start coding
- No CLI, no account, no config files
- Code executes in real Cloudflare Workers runtime

**Instant Preview:**
- Live preview pane with browser tab or HTTP tester
- Auto-reload on code changes
- DevTools integration (right-click → Inspect)

**Share & Deploy:**
- Copy Link generates permanent shareable URL
- Deploy button publishes to production in ~30 seconds
- Get `*.workers.dev` subdomain immediately

## Common Use Cases

- **API development:** Test endpoints before wrangler setup
- **Learning Workers:** Experiment with APIs without local environment
- **Prototyping:** Quick POCs for edge logic
- **Sharing examples:** Generate shareable links for bug reports or demos
- **Framework testing:** Import from CDN (Hono, itty-router, etc.)

## Limitations vs Production

| Feature | Playground | Production (wrangler) |
|---------|------------|----------------------|
| Language | JavaScript only | JS + TypeScript |
| Bindings | None | KV, D1, R2, DO, AI, etc. |
| Environment vars | None | Full support |
| Module format | ES only | ES + Service Worker |
| CPU time | 10ms (Free plan) | 10ms Free / 30s default, 5min max Paid |
| Custom domains | No | Yes |
| Analytics | No | Yes |

## See Also

- [Cloudflare Workers Docs](https://developers.cloudflare.com/workers/)
- [Workers Examples](https://developers.cloudflare.com/workers/examples/)
- [Wrangler CLI](https://developers.cloudflare.com/workers/wrangler/)
- [Workers API Reference](https://developers.cloudflare.com/workers/runtime-apis/)

# TCP Sockets API Reference

Complete API reference for the Cloudflare Workers TCP Sockets API (`cloudflare:sockets`).

## Core Function: `connect()`

```typescript
function connect(
  address: SocketAddress,
  options?: SocketOptions
): Socket
```

Creates an outbound TCP connection to the specified address.

### Parameters

#### `SocketAddress`

```typescript
interface SocketAddress {
  hostname: string; // DNS hostname or IP address
  port: number;     // TCP port (1-65535, excluding blocked ports)
}
```

| Field | Type | Description | Example |
|-------|------|-------------|---------|
| `hostname` | `string` | Target hostname or IP | `"db.internal.net"`, `"10.0.1.50"` |
| `port` | `number` | TCP port number | `5432`, `443`, `22` |

DNS names are resolved at connection time. IPv4, IPv6, and private IPs (10.x, 172.16.x, 192.168.x) supported.

#### `SocketOptions`

```typescript
interface SocketOptions {
  secureTransport?: "off" | "on" | "starttls";
  allowHalfOpen?: boolean;
}
```

| Field | Type | Default | Description |
|-------|------|---------|-------------|
| `secureTransport` | `"off" \| "on" \| "starttls"` | `"off"` | TLS mode |
| `allowHalfOpen` | `boolean` | `false` | Allow half-closed connections |

**`secureTransport` modes:**

| Mode | Behavior | Use Case |
|------|----------|----------|
| `"off"` | Plain TCP, no encryption | Testing, internal trusted networks |
| `"on"` | Immediate TLS handshake | HTTPS, secure databases, SSH |
| `"starttls"` | Start plain, upgrade later with `startTls()` | Postgres, SMTP, IMAP |

**`allowHalfOpen`:** When `false` (default), closing read stream auto-closes write stream. When `true`, streams are independent.

### Returns

A `Socket` object with readable/writable streams.

## Socket Interface

```typescript
interface Socket {
  // Streams
  readable: ReadableStream<Uint8Array>;
  writable: WritableStream<Uint8Array>;
  
  // Connection state
  opened: Promise<SocketInfo>;
  closed: Promise<void>;
  
  // Methods
  close(): Promise<void>;
  startTls(): Socket;
}
```

### Properties

#### `readable: ReadableStream<Uint8Array>`

Stream for reading data from the socket. Use `getReader()` to consume data.

```typescript
const reader = socket.readable.getReader();
const { done, value } = await reader.read(); // Read one chunk
```

#### `writable: WritableStream<Uint8Array>`

Stream for writing data to the socket. Use `getWriter()` to send data.

```typescript
const writer = socket.writable.getWriter();
await writer.write(new TextEncoder().encode("HELLO\r\n"));
await writer.close();
```

#### `opened: Promise<SocketInfo>`

Promise that resolves when connection succeeds, rejects on failure.

```typescript
interface SocketInfo {
  remoteAddress?: string; // May be undefined
  localAddress?: string;  // May be undefined
}

try {
  const info = await socket.opened;
} catch (error) {
  // Connection failed
}
```

#### `closed: Promise<void>`

Promise that resolves when socket is fully closed (both directions).

### Methods

#### `close(): Promise<void>`

Closes the socket gracefully, waiting for pending writes to complete.

```typescript
const socket = connect({ hostname: "api.internal", port: 443 });
try {
  // Use socket
} finally {
  await socket.close(); // Always call in finally block
}
```

#### `startTls(): Socket`

Upgrades connection to TLS. Only available when `secureTransport: "starttls"` was specified.

```typescript
const socket = connect(
  { hostname: "db.internal", port: 5432 },
  { secureTransport: "starttls" }
);

// Send protocol-specific StartTLS command
const writer = socket.writable.getWriter();
await writer.write(new TextEncoder().encode("STARTTLS\r\n"));

// Upgrade to TLS - use returned socket, not original
const secureSocket = socket.startTls();
const secureWriter = secureSocket.writable.getWriter();
```

## Complete Example

```typescript
import { connect } from 'cloudflare:sockets';

export default {
  async fetch(req: Request): Promise<Response> {
    const socket = connect({ hostname: "echo.example.com", port: 7 }, { secureTransport: "on" });

    try {
      await socket.opened;
      
      const writer = socket.writable.getWriter();
      await writer.write(new TextEncoder().encode("Hello, TCP!\n"));
      await writer.close();

      const reader = socket.readable.getReader();
      const { value } = await reader.read();
      
      return new Response(value);
    } finally {
      await socket.close();
    }
  }
};
```

See [patterns.md](./patterns.md) for multi-chunk reading, error handling, and protocol implementations.

## Quick Reference

| Task | Code |
|------|------|
| Import | `import { connect } from 'cloudflare:sockets';` |
| Connect | `connect({ hostname: "host", port: 443 })` |
| With TLS | `connect(addr, { secureTransport: "on" })` |
| StartTLS | `socket.startTls()` after handshake |
| Write | `await writer.write(data); await writer.close();` |
| Read | `const { value } = await reader.read();` |
| Error handling | `try { await socket.opened; } catch { }` |
| Always close | `try { } finally { await socket.close(); }` |

## See Also

- [patterns.md](./patterns.md) - Real-world protocol implementations
- [configuration.md](./configuration.md) - Wrangler setup and environment variables
- [gotchas.md](./gotchas.md) - Limits and error handling

# Configuration

Setup and configuration for TCP Sockets in Cloudflare Workers.

## Wrangler Configuration

### Basic Setup

TCP Sockets are available by default in Workers runtime. No special configuration required in `wrangler.jsonc`:

```jsonc
{
  "name": "private-network-worker",
  "main": "src/index.ts",
  "compatibility_date": "2025-01-01"
}
```

### Environment Variables

Store connection details as env vars:

```jsonc
{
  "vars": { "DB_HOST": "10.0.1.50", "DB_PORT": "5432" }
}
```

```typescript
interface Env { DB_HOST: string; DB_PORT: string; }

export default {
  async fetch(req: Request, env: Env): Promise<Response> {
    const socket = connect({ hostname: env.DB_HOST, port: parseInt(env.DB_PORT) });
  }
};
```

### Per-Environment Configuration

```jsonc
{
  "vars": { "DB_HOST": "localhost" },
  "env": {
    "staging": { "vars": { "DB_HOST": "staging-db.internal.net" } },
    "production": { "vars": { "DB_HOST": "prod-db.internal.net" } }
  }
}
```

Deploy: `wrangler deploy --env staging` or `wrangler deploy --env production`

## Integration with Cloudflare Tunnel

To connect Workers to private networks, combine TCP Sockets with Cloudflare Tunnel:

```
Worker (TCP Socket) → Tunnel hostname → cloudflared → Private Network
```

### Quick Setup

1. **Install cloudflared** on a server inside your private network
2. **Create tunnel**: `cloudflared tunnel create my-private-network`
3. **Configure routing** in `config.yml`:

```yaml
tunnel: <TUNNEL_ID>
credentials-file: /path/to/<TUNNEL_ID>.json
ingress:
  - hostname: db.internal.example.com
    service: tcp://10.0.1.50:5432
  - service: http_status:404  # Required catch-all
```

4. **Run tunnel**: `cloudflared tunnel run my-private-network`
5. **Connect from Worker**:

```typescript
const socket = connect(
  { hostname: "db.internal.example.com", port: 5432 },  // Tunnel hostname
  { secureTransport: "on" }
);
```

For detailed Tunnel setup, see [Tunnel configuration reference](../tunnel/configuration.md).

## Smart Placement Integration

Reduce latency by auto-placing Workers near backends:

```jsonc
{ "placement": { "mode": "smart" } }
```

Workers automatically relocate closer to TCP socket destinations after observing connection latency. See [Smart Placement reference](../smart-placement/).

## Secrets Management

Store sensitive credentials as secrets (not in wrangler.jsonc):

```bash
wrangler secret put DB_PASSWORD  # Enter value when prompted
```

Access in Worker via `env.DB_PASSWORD`. Use in protocol handshake or authentication.

## Local Development

Test with `wrangler dev`. Note: Local mode may not access private networks. Use public endpoints or mock servers for development:

```typescript
const config = process.env.NODE_ENV === 'dev' 
  ? { hostname: 'localhost', port: 5432 }  // Mock
  : { hostname: 'db.internal.example.com', port: 5432 };  // Production
```

## Connection String Patterns

Parse connection strings to extract host and port:

```typescript
function parseConnectionString(connStr: string): SocketAddress {
  const url = new URL(connStr); // e.g., "postgres://10.0.1.50:5432/mydb"
  return { hostname: url.hostname, port: parseInt(url.port) || 5432 };
}
```

## Hyperdrive Integration

For PostgreSQL/MySQL, prefer Hyperdrive over raw TCP sockets (includes connection pooling):

```jsonc
{ "hyperdrive": [{ "binding": "DB", "id": "<HYPERDRIVE_ID>" }] }
```

See [Hyperdrive reference](../hyperdrive/) for complete setup.

## Compatibility

TCP Sockets available in all modern Workers. Use current date: `"compatibility_date": "2025-01-01"`. No special flags required.

## Related Configuration

- **[Tunnel Configuration](../tunnel/configuration.md)** - Detailed cloudflared setup
- **[Smart Placement](../smart-placement/configuration.md)** - Placement mode options
- **[Hyperdrive](../hyperdrive/configuration.md)** - Database connection pooling setup

# Gotchas and Troubleshooting

Common pitfalls, limitations, and solutions for TCP Sockets in Cloudflare Workers.

## Platform Limits

### Connection Limits

| Limit | Value |
|-------|-------|
| Max concurrent sockets per request | 6 (hard limit) |
| Socket lifetime | Request duration |
| Connection timeout | Platform-dependent, no setting |

**Problem:** Exceeding 6 connections throws error

**Solution:** Process in batches of 6

```typescript
for (let i = 0; i < hosts.length; i += 6) {
  const batch = hosts.slice(i, i + 6).map(h => connect({ hostname: h, port: 443 }));
  await Promise.all(batch.map(async s => { /* use */ await s.close(); }));
}
```

### Blocked Destinations

Cloudflare IPs (1.1.1.1), localhost (127.0.0.1), port 25 (SMTP), Worker's own URL blocked for security.

**Solution:** Use public IPs or Tunnel hostnames: `connect({ hostname: "db.internal.company.net", port: 5432 })`

### Scope Requirements

**Problem:** Sockets created in global scope fail

**Cause:** Sockets tied to request lifecycle

**Solution:** Create inside handler: `export default { async fetch() { const socket = connect(...); } }`

## Common Errors

### Error: "proxy request failed"

**Causes:** Blocked destination (Cloudflare IP, localhost, port 25), DNS failure, network unreachable

**Solution:** Validate destinations, use Tunnel hostnames, catch errors with try/catch

### Error: "TCP Loop detected"

**Cause:** Worker connecting to itself

**Solution:** Connect to external service, not Worker's own hostname

### Error: "Port 25 prohibited"

**Cause:** SMTP port blocked

**Solution:** Use Email Workers API for email

### Error: "socket is not open"

**Cause:** Read/write after close

**Solution:** Always use try/finally to ensure proper closure order

### Error: Connection timeout

**Cause:** No built-in timeout

**Solution:** Use `Promise.race()`:

```typescript
const socket = connect(addr, opts);
const timeout = new Promise((_, reject) => setTimeout(() => reject(new Error('Timeout')), 5000));
await Promise.race([socket.opened, timeout]);
```

## TLS/SSL Issues

### StartTLS Timing

**Problem:** Calling `startTls()` too early

**Solution:** Send protocol-specific STARTTLS command, wait for server OK, then call `socket.startTls()`

### Certificate Validation

**Problem:** Self-signed certs fail

**Solution:** Use proper certs or Tunnel (handles TLS termination)

## Performance Issues

### Not Using Connection Pooling

**Problem:** New connection overhead per request

**Solution:** Use [Hyperdrive](../hyperdrive/) for databases (built-in pooling)

### Not Using Smart Placement

**Problem:** High latency to backend

**Solution:** Enable: `{ "placement": { "mode": "smart" } }` in wrangler.jsonc

### Forgetting to Close Sockets

**Problem:** Resource leaks

**Solution:** Always use try/finally:

```typescript
const socket = connect({ hostname: "api.internal", port: 443 });
try {
  // Use socket
} finally {
  await socket.close();
}
```

## Data Handling Issues

### Assuming Single Read Gets All Data

**Problem:** Only reading once may miss chunked data

**Solution:** Loop `reader.read()` until `done === true` (see patterns.md)

### Text Encoding Issues

**Problem:** Using wrong encoding

**Solution:** Specify encoding: `new TextDecoder('iso-8859-1').decode(data)`

## Security Issues

### SSRF Vulnerability

**Problem:** User-controlled destinations allow access to internal services

**Solution:** Validate against strict allowlist:

```typescript
const ALLOWED = ['api1.internal.net', 'api2.internal.net'];
const host = new URL(req.url).searchParams.get('host');
if (!host || !ALLOWED.includes(host)) return new Response('Forbidden', { status: 403 });
```

## When to Use Alternatives

| Use Case | Alternative | Reason |
|----------|-------------|--------|
| PostgreSQL/MySQL | [Hyperdrive](../hyperdrive/) | Connection pooling, caching |
| HTTP/HTTPS | `fetch()` | Simpler, built-in |
| HTTP with SSRF protection | VPC Services (beta 2025+) | Declarative bindings |

## Debugging Tips

1. **Log connection details:** `const info = await socket.opened; console.log(info.remoteAddress);`
2. **Test with public services first:** Use tcpbin.com:4242 echo server
3. **Verify Tunnel:** `cloudflared tunnel info <name>` and `cloudflared tunnel route ip list`

## Related

- [Hyperdrive](../hyperdrive/) - Database connections
- [Smart Placement](../smart-placement/) - Latency optimization
- [Tunnel Troubleshooting](../tunnel/gotchas.md)

# Common Patterns

Real-world patterns and examples for TCP Sockets in Cloudflare Workers.

```typescript
import { connect } from 'cloudflare:sockets';
```

## Basic Patterns

### Simple Request-Response

```typescript
const socket = connect({ hostname: "echo.example.com", port: 7 }, { secureTransport: "on" });
try {
  await socket.opened;
  const writer = socket.writable.getWriter();
  await writer.write(new TextEncoder().encode("Hello\n"));
  await writer.close();
  
  const reader = socket.readable.getReader();
  const { value } = await reader.read();
  return new Response(value);
} finally {
  await socket.close();
}
```

### Reading All Data

```typescript
async function readAll(socket: Socket): Promise<Uint8Array> {
  const reader = socket.readable.getReader();
  const chunks: Uint8Array[] = [];
  while (true) {
    const { done, value } = await reader.read();
    if (done) break;
    chunks.push(value);
  }
  const total = chunks.reduce((sum, c) => sum + c.length, 0);
  const result = new Uint8Array(total);
  let offset = 0;
  for (const chunk of chunks) { result.set(chunk, offset); offset += chunk.length; }
  return result;
}
```

### Streaming Response

```typescript
// Stream socket data directly to HTTP response
const socket = connect({ hostname: "stream.internal", port: 9000 }, { secureTransport: "on" });
const writer = socket.writable.getWriter();
await writer.write(new TextEncoder().encode("STREAM\n"));
await writer.close();
return new Response(socket.readable);
```

## Protocol Examples

### Redis RESP

```typescript
// Send: *2\r\n$3\r\nGET\r\n$<keylen>\r\n<key>\r\n
// Recv: $<len>\r\n<data>\r\n or $-1\r\n for null
const socket = connect({ hostname: "redis.internal", port: 6379 });
const writer = socket.writable.getWriter();
await writer.write(new TextEncoder().encode(`*2\r\n$3\r\nGET\r\n$3\r\nkey\r\n`));
```

### PostgreSQL

**Use [Hyperdrive](../hyperdrive/) for production.** Raw Postgres protocol is complex (startup, auth, query messages).

### MQTT

```typescript
const socket = connect({ hostname: "mqtt.broker", port: 1883 });
const writer = socket.writable.getWriter();
// CONNECT: 0x10 <len> 0x00 0x04 "MQTT" 0x04 <flags> ...
// PUBLISH: 0x30 <len> <topic_len> <topic> <message>
```

## Error Handling Patterns

### Retry with Backoff

```typescript
async function connectWithRetry(addr: SocketAddress, opts: SocketOptions, maxRetries = 3): Promise<Socket> {
  for (let i = 1; i <= maxRetries; i++) {
    try {
      const socket = connect(addr, opts);
      await socket.opened;
      return socket;
    } catch (error) {
      if (i === maxRetries) throw error;
      await new Promise(r => setTimeout(r, 1000 * Math.pow(2, i - 1))); // Exponential backoff
    }
  }
  throw new Error('Unreachable');
}
```

### Timeout

```typescript
async function connectWithTimeout(addr: SocketAddress, opts: SocketOptions, ms = 5000): Promise<Socket> {
  const socket = connect(addr, opts);
  const timeout = new Promise<never>((_, reject) => setTimeout(() => reject(new Error('Timeout')), ms));
  await Promise.race([socket.opened, timeout]);
  return socket;
}
```

### Fallback

```typescript
async function connectWithFallback(primary: string, fallback: string, port: number): Promise<Socket> {
  try {
    const socket = connect({ hostname: primary, port }, { secureTransport: "on" });
    await socket.opened;
    return socket;
  } catch {
    return connect({ hostname: fallback, port }, { secureTransport: "on" });
  }
}
```

## Security Patterns

### Destination Allowlist (Prevent SSRF)

```typescript
const ALLOWED_HOSTS = ['db.internal.company.net', 'api.internal.company.net', /^10\.0\.1\.\d+$/];

function isAllowed(hostname: string): boolean {
  return ALLOWED_HOSTS.some(p => p instanceof RegExp ? p.test(hostname) : p === hostname);
}

export default {
  async fetch(req: Request): Promise<Response> {
    const target = new URL(req.url).searchParams.get('host');
    if (!target || !isAllowed(target)) return new Response('Forbidden', { status: 403 });
    const socket = connect({ hostname: target, port: 443 });
    // Use socket...
  }
};
```

### Connection Pooling

```typescript
class SocketPool {
  private pool = new Map<string, Socket[]>();
  
  async acquire(hostname: string, port: number): Promise<Socket> {
    const key = `${hostname}:${port}`;
    const sockets = this.pool.get(key) || [];
    if (sockets.length > 0) return sockets.pop()!;
    const socket = connect({ hostname, port }, { secureTransport: "on" });
    await socket.opened;
    return socket;
  }
  
  release(hostname: string, port: number, socket: Socket): void {
    const key = `${hostname}:${port}`;
    const sockets = this.pool.get(key) || [];
    if (sockets.length < 3) { sockets.push(socket); this.pool.set(key, sockets); }
    else socket.close();
  }
}
```

## Multi-Protocol Gateway

```typescript
interface Protocol { name: string; defaultPort: number; test(host: string, port: number): Promise<string>; }

const PROTOCOLS: Record<string, Protocol> = {
  redis: {
    name: 'redis',
    defaultPort: 6379,
    async test(host, port) {
      const socket = connect({ hostname: host, port });
      try {
        const writer = socket.writable.getWriter();
        await writer.write(new TextEncoder().encode('*1\r\n$4\r\nPING\r\n'));
        writer.releaseLock();
        const reader = socket.readable.getReader();
        const { value } = await reader.read();
        return new TextDecoder().decode(value || new Uint8Array());
      } finally { await socket.close(); }
    }
  }
};

export default {
  async fetch(req: Request): Promise<Response> {
    const url = new URL(req.url);
    const proto = url.pathname.slice(1);  // /redis
    const host = url.searchParams.get('host');
    if (!host || !PROTOCOLS[proto]) return new Response('Invalid', { status: 400 });
    const result = await PROTOCOLS[proto].test(host, parseInt(url.searchParams.get('port') || '') || PROTOCOLS[proto].defaultPort);
    return new Response(result);
  }
};
```

# Workers VPC Connectivity

Connect Cloudflare Workers to private networks and internal infrastructure using TCP Sockets.

## Overview

Workers VPC connectivity enables outbound TCP connections from Workers to private resources in AWS, Azure, GCP, on-premises datacenters, or any private network. This is achieved through the **TCP Sockets API** (`cloudflare:sockets`), which provides low-level network access for custom protocols and services.

**Key capabilities:**
- Direct TCP connections to private IPs and hostnames
- TLS/StartTLS support for encrypted connections
- Integration with Cloudflare Tunnel for secure private network access
- Full control over wire protocols (database protocols, SSH, MQTT, custom TCP)

**Note:** This reference documents the TCP Sockets API. For the newer Workers VPC Services product (HTTP-only service bindings with built-in SSRF protection), refer to separate documentation when available. VPC Services is currently in beta (2025+).

## Quick Decision: Which Technology?

Need private network connectivity from Workers?

| Requirement | Use | Why |
|------------|-----|-----|
| HTTP/HTTPS APIs in private network | VPC Services (beta, separate docs) | SSRF-safe, declarative bindings |
| PostgreSQL/MySQL databases | [Hyperdrive](../hyperdrive/) | Connection pooling, caching, optimized |
| Custom TCP protocols (SSH, MQTT, proprietary) | **TCP Sockets (this doc)** | Full protocol control |
| Simple HTTP with lowest latency | TCP Sockets + [Smart Placement](../smart-placement/) | Manual optimization |
| Expose on-prem to internet (inbound) | [Cloudflare Tunnel](../tunnel/) | Not Worker-specific |

## When to Use TCP Sockets

**Use TCP Sockets when you need:**
- ✅ Direct control over wire protocols (e.g., Postgres wire protocol, SSH, Redis RESP)
- ✅ Non-HTTP protocols (MQTT, SMTP, custom binary protocols)
- ✅ StartTLS or custom TLS negotiation
- ✅ Streaming binary data over TCP

**Don't use TCP Sockets when:**
- ❌ You just need HTTP/HTTPS (use `fetch()` or VPC Services)
- ❌ You need PostgreSQL/MySQL (use Hyperdrive for pooling)
- ❌ You need WebSocket (use native Workers WebSocket)

## Quick Start

```typescript
import { connect } from 'cloudflare:sockets';

export default {
  async fetch(req: Request): Promise<Response> {
    // Connect to private service
    const socket = connect(
      { hostname: "db.internal.company.net", port: 5432 },
      { secureTransport: "on" }
    );

    try {
      await socket.opened; // Wait for connection
      
      const writer = socket.writable.getWriter();
      await writer.write(new TextEncoder().encode("QUERY\r\n"));
      await writer.close();

      const reader = socket.readable.getReader();
      const { value } = await reader.read();
      
      return new Response(value);
    } finally {
      await socket.close();
    }
  }
};
```

## Architecture Pattern: Workers + Tunnel

Most private network connectivity combines TCP Sockets with Cloudflare Tunnel:

```
┌─────────┐     ┌─────────────┐     ┌──────────────┐     ┌─────────────┐
│ Worker  │────▶│ TCP Socket  │────▶│   Tunnel     │────▶│   Private   │
│         │     │ (this API)  │     │ (cloudflared)│     │   Network   │
└─────────┘     └─────────────┘     └──────────────┘     └─────────────┘
```

1. Worker opens TCP socket to Tunnel hostname
2. Tunnel endpoint routes to private IP
3. Response flows back through Tunnel to Worker

See [configuration.md](./configuration.md) for Tunnel setup details.

## Reading Order

1. **Start here (README.md)** - Overview and decision guide
2. **[api.md](./api.md)** - Socket interface, types, methods
3. **[configuration.md](./configuration.md)** - Wrangler setup, Tunnel integration
4. **[patterns.md](./patterns.md)** - Real-world examples (databases, protocols, error handling)
5. **[gotchas.md](./gotchas.md)** - Limits, blocked ports, common errors

## Key Limits

| Limit | Value |
|-------|-------|
| Max concurrent sockets per request | 6 |
| Blocked destinations | Cloudflare IPs, localhost, port 25 |
| Scope requirement | Must create in handler (not global) |

See [gotchas.md](./gotchas.md) for complete limits and troubleshooting.

## Best Practices

1. **Always close sockets** - Use try/finally blocks
2. **Validate destinations** - Prevent SSRF by allowlisting hosts
3. **Use Hyperdrive for databases** - Better performance than raw TCP
4. **Prefer fetch() for HTTP** - Only use TCP when necessary
5. **Combine with Smart Placement** - Reduce latency to private networks

## Related Technologies

- **[Hyperdrive](../hyperdrive/)** - PostgreSQL/MySQL with connection pooling
- **[Cloudflare Tunnel](../tunnel/)** - Secure private network access
- **[Smart Placement](../smart-placement/)** - Auto-locate Workers near backends
- **VPC Services (beta)** - HTTP-only service bindings with SSRF protection (separate docs)

## Reference

- [TCP Sockets API Documentation](https://developers.cloudflare.com/workers/runtime-apis/tcp-sockets/)
- [Connect to databases guide](https://developers.cloudflare.com/workers/tutorials/postgres/)
- [Cloudflare Tunnel setup](https://developers.cloudflare.com/cloudflare-one/connections/connect-networks/)

# Workflow APIs

Fetch the documentation for the operation before writing code; use its current signatures and serialization rules.

| Task | Documentation |
| --- | --- |
| Implement steps; create, batch, inspect, pause, resume, restart, or terminate instances; check parameter and return types | [Workers API](https://developers.cloudflare.com/workflows/build/workers-api/) |
| Access a step's name, occurrence, retry attempt, and resolved configuration | [Step context](https://developers.cloudflare.com/workflows/build/step-context/) |
| Configure retries, backoff, timeouts, non-retryable failures, or relative/absolute sleeps | [Sleeping and retrying](https://developers.cloudflare.com/workflows/build/sleeping-and-retrying/) |
| Pass initial parameters, wait for an external event, or send an event to an instance | [Events and parameters](https://developers.cloudflare.com/workflows/build/events-and-parameters/) |
| Start instances from a Worker or another Workflow, or schedule execution | [Trigger Workflows](https://developers.cloudflare.com/workflows/build/trigger-workflows/) |
| Trigger and manage instances from the command line | [Wrangler commands](https://developers.cloudflare.com/workflows/reference/wrangler-commands/) |
| Manage Workflows over HTTP, including authentication and request bodies | [Workflows REST API](https://developers.cloudflare.com/api/resources/workflows/methods/list/) |

Decide whether input is available at creation or must arrive later as an event. Starting a child instance does not establish that it has completed; choose how the parent will observe completion. Check current instance ID retention and creation semantics before designing duplicate-trigger handling.

See [configuration.md](./configuration.md) and [patterns.md](./patterns.md).

# Workflow Configuration

Use the current guides for configuration fields and setup commands; check the project's installed Wrangler version and generated binding types before adapting an existing project.

| Task | Documentation |
| --- | --- |
| Scaffold and deploy a Workflow class with its binding | [Build your first Workflow](https://developers.cloudflare.com/workflows/get-started/guide/) |
| Configure one or more Workflows, including a binding to a Workflow in another Worker | [Wrangler Workflows configuration](https://developers.cloudflare.com/workers/wrangler/configuration/#workflows) |
| Configure storage, AI, and other resources used by steps | [Workers bindings](https://developers.cloudflare.com/workers/runtime-apis/bindings/) |
| Choose step retry, backoff, timeout, and sleep behavior | [Sleeping and retrying](https://developers.cloudflare.com/workflows/build/sleeping-and-retrying/) |
| Schedule instances or trigger them from another Worker | [Trigger Workflows](https://developers.cloudflare.com/workflows/build/trigger-workflows/) |
| Trigger a Workflow from Pages Functions | [Call Workflows from Pages](https://developers.cloudflare.com/workflows/build/call-workflows-from-pages/) |
| Develop and inspect instances locally | [Local development](https://developers.cloudflare.com/workflows/build/local-development/) |
| Set resource budgets and inspect execution | [Limits](https://developers.cloudflare.com/workflows/reference/limits/) and [metrics and analytics](https://developers.cloudflare.com/workflows/observability/metrics-analytics/) |

Distinguish the Worker that defines the Workflow from callers that trigger it. For Pages, follow the documented intermediary Worker/service-binding approach. A step's elapsed-time timeout and the Worker's active CPU budget address different failure modes; configure them based on the actual failure.

See [api.md](./api.md), [patterns.md](./patterns.md), and [gotchas.md](./gotchas.md).

# Gotchas & Debugging

Start with the failing instance and step, then fetch the relevant guide before changing code or resource limits.

| Symptom or question | What to check |
| --- | --- |
| Step timeout or repeated failure | [Sleeping and retrying](https://developers.cloudflare.com/workflows/build/sleeping-and-retrying/) for per-attempt timeout, retry policy, and non-retryable failures |
| CPU exhaustion despite a short run | [Limits](https://developers.cloudflare.com/workflows/reference/limits/) for active CPU budgets; increasing an elapsed-time timeout does not increase CPU capacity |
| Missing event or event timeout | [Events and parameters](https://developers.cloudflare.com/workflows/build/events-and-parameters/) for instance targeting, event type/payload requirements, and timeout handling |
| State disappears or branches change after resuming | [Rules of Workflows](https://developers.cloudflare.com/workflows/build/rules-of-workflows/) for persisted step returns, deterministic names and conditionals, and awaited operations |
| Duplicate charge, write, or notification | Review the destination's idempotency guarantees and [step design](./patterns.md#design-decisions); retries can repeat an external operation even when its previous attempt committed |
| Instance ID collision or unexpected batch result | [Workers API](https://developers.cloudflare.com/workflows/build/workers-api/) for creation semantics, plus [limits](https://developers.cloudflare.com/workflows/reference/limits/) for retention |
| Oversized results, queued instances, or missing historical data | [Limits](https://developers.cloudflare.com/workflows/reference/limits/) for return/event sizes, concurrency, creation rates, and retention; export required long-term results before expiry |
| Local-only failure or failing introspection test | [Local development](https://developers.cloudflare.com/workflows/build/local-development/) and [Workflow test APIs](https://developers.cloudflare.com/workers/testing/vitest-integration/test-apis/#workflows) |
| Inspect execution and cost | [Metrics and analytics](https://developers.cloudflare.com/workflows/observability/metrics-analytics/), [Wrangler commands](https://developers.cloudflare.com/workflows/reference/wrangler-commands/), and [pricing](https://developers.cloudflare.com/workflows/reference/pricing/) |

CPU time measures active computation; waiting for network or storage I/O is elapsed time. Event waits, sleeps, and retry delays also have their own documented behavior. Check the current limits page for how these states affect concurrency and step accounting rather than treating every wait as active execution.

See [README.md](./README.md), [configuration.md](./configuration.md), [api.md](./api.md), and [patterns.md](./patterns.md).

# Workflow Patterns

## Design Decisions

Read [Rules of Workflows](https://developers.cloudflare.com/workflows/build/rules-of-workflows/) before choosing step boundaries or concurrency patterns.

- Separate work into steps that can be retried independently. Persist results through step returns and keep side effects inside steps.
- Make side effects safe to repeat. A retry can happen after an external write succeeds; use the destination's idempotency mechanism or atomic deduplication. A separate check followed by a write does not itself guarantee idempotency.
- Base step names, loops, and branches on stable input or persisted results. In-memory state and fresh time/random values cannot serve as durable replay state.
- Await step operations, and check the documented replay behavior before combining steps in parallel or racing them.
- Keep large data in external storage when appropriate and pass references between steps; consult current return-type and size constraints.

## Examples and Orchestration

| Task | Documentation |
| --- | --- |
| Process images with human approval; handle approval events and timeouts | [Human-in-the-loop image tagging](https://developers.cloudflare.com/workflows/examples/wait-for-event/) |
| Implement a payment and notification sequence | [Pay cart and send invoice](https://developers.cloudflare.com/workflows/examples/send-invoices/) |
| Export data to object storage | [Export and save D1 database](https://developers.cloudflare.com/workflows/examples/backup-d1/) |
| Delay lifecycle follow-ups or retry transient failures | [Sleeping and retrying](https://developers.cloudflare.com/workflows/build/sleeping-and-retrying/) |
| Schedule jobs or start child Workflows | [Trigger Workflows](https://developers.cloudflare.com/workflows/build/trigger-workflows/) |
| Design parallel work, races, conditional steps, and batch creation | [Rules of Workflows](https://developers.cloudflare.com/workflows/build/rules-of-workflows/) |

## Testing Workflows

Fetch [Vitest setup](https://developers.cloudflare.com/workers/testing/vitest-integration/write-your-first-test/) for current dependencies and configuration, then use the [Workflow test APIs](https://developers.cloudflare.com/workers/testing/vitest-integration/test-apis/#workflows) for introspection, step/event mocks, sleep controls, and cleanup.

Test retry behavior, event arrival and timeout paths, and duplicate external effects. Use documented introspection waits to observe completion rather than assuming a newly created instance has finished.

See [configuration.md](./configuration.md), [api.md](./api.md), and [gotchas.md](./gotchas.md).

# Cloudflare Workflows

Use Workflows for durable, multi-step jobs that must retry, wait, and resume without losing completed work. An instance is one execution; steps define persistence and retry boundaries.

Fetch the relevant current documentation before implementing. API shapes, configuration, testing helpers, limits, and examples belong in the docs rather than in this reference.

- **Start a project:** [Build your first Workflow](https://developers.cloudflare.com/workflows/get-started/guide/) covers scaffolding, configuration, deployment, and a first instance.
- **Design durable execution:** [Rules of Workflows](https://developers.cloudflare.com/workflows/build/rules-of-workflows/) covers step boundaries, replay, state, and idempotency.
- **Implement or manage an instance:** [Workers API](https://developers.cloudflare.com/workflows/build/workers-api/) covers steps, instance operations, parameters, and return types.
- **Check capacity and cost:** fetch [limits](https://developers.cloudflare.com/workflows/reference/limits/) and [pricing](https://developers.cloudflare.com/workflows/reference/pricing/) for the target plan.

## In This Reference

- [configuration.md](./configuration.md) — setup, bindings, retry configuration, and local development
- [api.md](./api.md) — steps, instance lifecycle, events, CLI, and REST operations
- [patterns.md](./patterns.md) — design decisions, examples, orchestration, and tests
- [gotchas.md](./gotchas.md) — failures, timeouts, replay, and capacity investigation

## See Also

- [Durable Objects](https://developers.cloudflare.com/durable-objects/) — stateful coordination
- [Queues](../queues/README.md) — asynchronous message delivery
- [Workers](https://developers.cloudflare.com/workers/) — application entry points that trigger instances

# Zaraz Web API

Client-side JavaScript API for tracking events, setting properties, and managing consent.

## zaraz.track()

```javascript
zaraz.track('button_click');
zaraz.track('purchase', { value: 99.99, currency: 'USD', item_id: '12345' });
zaraz.track('pageview', { page_path: '/products', page_title: 'Products' }); // SPA
```

**Params:** `eventName` (string), `properties` (object, optional). Fire-and-forget.

## zaraz.set()

```javascript
zaraz.set('userId', 'user_12345');
zaraz.set({ email: '[email protected]', plan: 'premium', country: 'US' });
```

Properties persist for page session. Use for user identification and segmentation.

## zaraz.ecommerce()

```javascript
zaraz.ecommerce('Product Viewed', { product_id: 'SKU123', name: 'Widget', price: 49.99 });
zaraz.ecommerce('Product Added', { product_id: 'SKU123', quantity: 2, price: 49.99 });
zaraz.ecommerce('Order Completed', {
  order_id: 'ORD-789', total: 149.98, currency: 'USD',
  products: [{ product_id: 'SKU123', quantity: 2, price: 49.99 }]
});
```

**Events:** `Product Viewed`, `Product Added`, `Product Removed`, `Cart Viewed`, `Checkout Started`, `Order Completed`

Tools auto-map to GA4, Facebook CAPI, etc.

## System Properties (Triggers)

```
{{system.page.url}}   {{system.page.title}}   {{system.page.referrer}}
{{system.device.ip}}  {{system.device.userAgent}}  {{system.device.language}}
{{system.cookies.name}}  {{client.__zarazTrack.userId}}
```

## zaraz.consent

```javascript
// Check
const purposes = zaraz.consent.getAll(); // { analytics: true, marketing: false }

// Set
zaraz.consent.modal = true; // Show modal
zaraz.consent.setAll({ analytics: true, marketing: false });
zaraz.consent.set('marketing', true);

// Listen
zaraz.consent.addEventListener('consentChanged', () => {
  if (zaraz.consent.getAll().marketing) zaraz.track('marketing_consent_granted');
});
```

**Flow:** Configure purposes in dashboard → Map tools to purposes → Show modal/set programmatically → Tools fire when allowed

## zaraz.debug

```javascript
zaraz.debug = true;
zaraz.track('test_event');
console.log(zaraz.tools); // View loaded tools
```

## Cookie Methods

```javascript
zaraz.getCookie('session_id');  // Zaraz namespace
zaraz.readCookie('_ga');        // Any cookie
```

## Async Behavior

All methods fire-and-forget. Events batched and sent asynchronously:

```javascript
zaraz.track('event1');
zaraz.set('prop', 'value');
zaraz.track('event2'); // All batched
```

## TypeScript Types

```typescript
interface Zaraz {
  track(event: string, properties?: Record<string, unknown>): void;
  set(key: string, value: unknown): void;
  set(properties: Record<string, unknown>): void;
  ecommerce(event: string, properties: Record<string, unknown>): void;
  consent: {
    getAll(): Record<string, boolean>;
    setAll(purposes: Record<string, boolean>): void;
    set(purpose: string, value: boolean): void;
    addEventListener(event: 'consentChanged', callback: () => void): void;
    modal: boolean;
  };
  debug: boolean;
  tools?: string[];
  getCookie(name: string): string | undefined;
  readCookie(name: string): string | undefined;
}
declare global { interface Window { zaraz: Zaraz; } }
```

# Zaraz Configuration

## Dashboard Setup

1. Domain → Zaraz → Start setup
2. Add tool (e.g., Google Analytics 4)
3. Enter credentials (GA4: `G-XXXXXXXXXX`)
4. Configure triggers
5. Save and Publish

## Triggers

| Type | When | Use Case |
|------|------|----------|
| Pageview | Page load | Track page views |
| Click | Element clicked | Button tracking |
| Form Submission | Form submitted | Lead capture |
| History Change | URL changes (SPA) | React/Vue routing |
| Variable Match | Custom condition | Conditional firing |

### History Change (SPA)

```
Type: History Change
Event: pageview
```

Fires on `pushState`, `replaceState`, hash changes. **No manual tracking needed.**

### Click Trigger

```
Type: Click
CSS Selector: .buy-button
Event: purchase_intent
Properties:
  button_text: {{system.clickElement.text}}
```

## Tool Configuration

**GA4:**
```
Measurement ID: G-XXXXXXXXXX
Events: page_view, purchase, user_engagement
```

**Facebook Pixel:**
```
Pixel ID: 1234567890123456
Events: PageView, Purchase, AddToCart
```

**Google Ads:**
```
Conversion ID: AW-XXXXXXXXX
Conversion Label: YYYYYYYYYY
```

## Consent Management

1. Settings → Consent → Create purposes (analytics, marketing)
2. Map tools to purposes
3. Set behavior: "Do not load until consent granted"

**Programmatic consent:**
```javascript
zaraz.consent.setAll({ analytics: true, marketing: true });
```

## Privacy Features

| Feature | Default |
|---------|---------|
| IP Anonymization | Enabled |
| Cookie Control | Via consent purposes |
| GDPR/CCPA | Consent modal |

## Testing

1. **Preview Mode** - test without publishing
2. **Debug Mode** - `zaraz.debug = true`
3. **Network tab** - filter "zaraz"

## Limits

| Resource | Limit |
|----------|-------|
| Event properties | 100KB |
| Consent purposes | 20 |

# Zaraz Gotchas

## Events Not Firing

**Check:**
1. Tool enabled in dashboard (green dot)
2. Trigger conditions met
3. Consent granted for tool's purpose
4. Tool credentials correct (GA4: `G-XXXXXXXXXX`, FB: numeric only)

**Debug:**
```javascript
zaraz.debug = true;
console.log('Tools:', zaraz.tools);
console.log('Consent:', zaraz.consent.getAll());
```

## Consent Issues

**Modal not showing:**
```javascript
// Clear consent cookie
document.cookie = 'zaraz-consent=; expires=Thu, 01 Jan 1970 00:00:00 UTC; path=/;';
location.reload();
```

**Tools firing before consent:** Map tool to consent purpose with "Do not load until consent granted".

## SPA Tracking

**Route changes not tracked:**
1. Configure History Change trigger in dashboard
2. Hash routing (`#/path`) requires manual tracking:
```javascript
window.addEventListener('hashchange', () => {
  zaraz.track('pageview', { page_path: location.pathname + location.hash });
});
```

**React fix:**
```javascript
const location = useLocation();
useEffect(() => {
  zaraz.track('pageview', { page_path: location.pathname });
}, [location]); // Include dependency
```

## Performance

**Slow page load:**
- Audit tool count (50+ degrades performance)
- Disable blocking triggers unless required
- Reduce event payload size (<100KB)

## Tool-Specific Issues

| Tool | Issue | Fix |
|------|-------|-----|
| GA4 | Events not in real-time | Wait 5-10 min, use DebugView |
| Facebook | Invalid Pixel ID | Use numeric only (no `fbpx_` prefix) |
| Google Ads | Conversions not attributed | Include `send_to: 'AW-XXX/LABEL'` |

## Data Layer

- Properties persist per page only - set on each page load
- Nested access: `{{client.__zarazTrack.user.plan}}`

## Limits

| Resource | Limit |
|----------|-------|
| Request size | 100KB |
| Consent purposes | 20 |
| API rate | 1000 req/sec |

## When NOT to Use Zaraz

- Server-to-server tracking (use Workers)
- Real-time bidirectional communication
- Binary data transmission
- Authentication flows

# Zaraz Reference Implementation Summary

## Files Created

| File | Lines | Purpose |
|------|-------|---------|
| README.md | 111 | Navigation, decision tree, quick start |
| api.md | 287 | Web API reference, Zaraz Context |
| configuration.md | 307 | Dashboard setup, triggers, tools, consent |
| patterns.md | 430 | SPA, e-commerce, Worker integration |
| gotchas.md | 317 | Troubleshooting, limits, tool-specific issues |
| **Total** | **1,452** | **vs 366 original** |

## Key Improvements Applied

### Structure
- ✅ Created 5-file progressive disclosure system
- ✅ Added navigation table in README
- ✅ Added decision tree for routing
- ✅ Added "Reading Order by Task" guide
- ✅ Cross-referenced files throughout

### New Content Added
- ✅ Zaraz Context (system/client properties)
- ✅ History Change trigger for SPA tracking
- ✅ Context Enrichers pattern
- ✅ Worker Variables pattern
- ✅ Consent management deep dive
- ✅ Tool-specific quirks (GA4, Facebook, Google Ads)
- ✅ GTM migration guide
- ✅ Comprehensive troubleshooting
- ✅ "When NOT to use Zaraz" section
- ✅ TypeScript type definitions

### Preserved Content
- ✅ All original API methods
- ✅ E-commerce tracking examples
- ✅ Consent management
- ✅ Workers integration (expanded)
- ✅ Common patterns (expanded)
- ✅ Debugging tools
- ✅ Reference links

## Progressive Disclosure Impact

### Before (Monolithic)
All tasks loaded 366 lines regardless of need.

### After (Progressive)
- **Track event task**: README (111) + api.md (287) = 398 lines
- **Debug issue**: gotchas.md (317) = 317 lines (13% reduction)
- **Configure tool**: configuration.md (307) = 307 lines (16% reduction)
- **SPA tracking**: README + patterns.md (SPA section) ~180 lines (51% reduction)

**Net effect:** Task-specific loading reduces unnecessary content by 13-51% depending on use case.

## File Summary

### README.md (111 lines)
- Overview and core concepts
- Quick start guide
- When to use Zaraz vs Workers
- Navigation table
- Reading order by task
- Decision tree

### api.md (287 lines)
- zaraz.track()
- zaraz.set()
- zaraz.ecommerce()
- Zaraz Context (system/client properties)
- zaraz.consent API
- zaraz.debug
- Cookie methods
- TypeScript definitions

### configuration.md (307 lines)
- Dashboard setup flow
- Trigger types (including History Change)
- Tool configuration (GA4, Facebook, Google Ads)
- Actions and action rules
- Selective loading
- Consent management setup
- Privacy features
- Testing workflow

### patterns.md (430 lines)
- SPA tracking (React, Vue, Next.js)
- User identification flows
- Complete e-commerce funnel
- A/B testing
- Worker integration (Context Enrichers, Worker Variables, HTML injection)
- Multi-tool coordination
- GTM migration
- Best practices

### gotchas.md (317 lines)
- Events not firing (5-step debug process)
- Consent issues
- SPA tracking pitfalls
- Performance issues
- Tool-specific quirks
- Data layer issues
- Limits table
- When NOT to use Zaraz
- Debug checklist

## Quality Metrics

- ✅ All files use consistent markdown formatting
- ✅ Code examples include language tags
- ✅ Tables for structured data (limits, parameters, comparisons)
- ✅ Problem → Cause → Solution format in gotchas
- ✅ Cross-references between files
- ✅ No "see documentation" placeholders
- ✅ Real, actionable examples throughout
- ✅ Verified API syntax for Workers

## Original Backup

Original SKILL.md preserved as `_SKILL_old.md` for reference.

# Zaraz Patterns

## SPA Tracking

**History Change Trigger (Recommended):** Configure in dashboard - no code needed, Zaraz auto-detects route changes.

**Manual tracking (React/Vue/Next.js):**
```javascript
// On route change
zaraz.track('pageview', { page_path: pathname, page_title: document.title });
```

## User Identification

```javascript
// Login
zaraz.set({ userId: user.id, email: user.email, plan: user.plan });
zaraz.track('login', { method: 'oauth' });

// Logout - set to null (cannot clear)
zaraz.set('userId', null);
```

## E-commerce Funnel

| Event | Method |
|-------|--------|
| View | `zaraz.ecommerce('Product Viewed', { product_id, name, price })` |
| Add to cart | `zaraz.ecommerce('Product Added', { product_id, quantity })` |
| Checkout | `zaraz.ecommerce('Checkout Started', { cart_id, products: [...] })` |
| Purchase | `zaraz.ecommerce('Order Completed', { order_id, total, products })` |

## A/B Testing

```javascript
zaraz.set('experiment_checkout', variant);
zaraz.track('experiment_viewed', { experiment_id: 'checkout', variant });
// On conversion
zaraz.track('experiment_conversion', { experiment_id, variant, value });
```

## Worker Integration

**Context Enricher** - Modify context before tools execute:
```typescript
export default {
  async fetch(request, env) {
    const body = await request.json();
    body.system.userRegion = request.cf?.region;
    return Response.json(body);
  }
};
```
Configure: Zaraz > Settings > Context Enrichers

**Worker Variables** - Compute dynamic values server-side, use as `{{worker.variable_name}}`.

## GTM Migration

| GTM | Zaraz |
|-----|-------|
| `dataLayer.push({event: 'purchase'})` | `zaraz.ecommerce('Order Completed', {...})` |
| `{{Page URL}}` | `{{system.page.url}}` |
| `{{Page Title}}` | `{{system.page.title}}` |
| Page View trigger | Pageview trigger |
| Click trigger | Click (selector: `*`) |

## Best Practices

1. Use dashboard triggers over inline code
2. Enable History Change for SPAs (no manual code)
3. Debug with `zaraz.debug = true`
4. Implement consent early (GDPR/CCPA)
5. Use Context Enrichers for sensitive/server data

# Cloudflare Zaraz

Expert guidance for Cloudflare Zaraz - server-side tag manager for loading third-party tools at the edge.

## What is Zaraz?

Zaraz offloads third-party scripts (analytics, ads, chat, marketing) to Cloudflare's edge, improving site speed, privacy, and security. Zero client-side performance impact.

**Core Concepts:**
- **Server-side execution** - Scripts run on Cloudflare, not user's browser
- **Single HTTP request** - All tools loaded via one endpoint
- **Privacy-first** - Control data sent to third parties
- **No client-side JS overhead** - Minimal browser impact

## Quick Start

1. Navigate to domain > Zaraz in Cloudflare dashboard
2. Click "Start setup"
3. Add tools (Google Analytics, Facebook Pixel, etc.)
4. Configure triggers (when tools fire)
5. Add tracking code to your site:

```javascript
// Track page view
zaraz.track('page_view');

// Track custom event
zaraz.track('button_click', { button_id: 'cta' });

// Set user properties
zaraz.set('userId', 'user_123');
```

## When to Use Zaraz

**Use Zaraz when:**
- Adding multiple third-party tools (analytics, ads, marketing)
- Site performance is critical (no client-side JS overhead)
- Privacy compliance required (GDPR, CCPA)
- Non-technical teams need to manage tools

**Use Workers directly when:**
- Building custom server-side tracking logic
- Need full control over data processing
- Integrating with complex backend systems
- Zaraz's tool library doesn't meet needs

## In This Reference

| File | Purpose | When to Read |
|------|---------|--------------|
| [api.md](./api.md) | Web API, zaraz object, consent methods | Implementing tracking calls |
| [configuration.md](./configuration.md) | Dashboard setup, triggers, tools | Initial setup, adding tools |
| [patterns.md](./patterns.md) | SPA, e-commerce, Worker integration | Best practices, common scenarios |
| [gotchas.md](./gotchas.md) | Troubleshooting, limits, pitfalls | Debugging issues |

## Reading Order by Task

| Task | Files to Read |
|------|---------------|
| Add analytics to site | README → configuration.md |
| Track custom events | README → api.md |
| Debug tracking issues | gotchas.md |
| SPA tracking | api.md → patterns.md (SPA section) |
| E-commerce tracking | api.md#ecommerce → patterns.md#ecommerce |
| Worker integration | patterns.md#worker-integration |
| GDPR compliance | api.md#consent → configuration.md#consent |

## Decision Tree

```
What do you need?

├─ Track events in browser → api.md
│   ├─ Page views, clicks → zaraz.track()
│   ├─ User properties → zaraz.set()
│   └─ E-commerce → zaraz.ecommerce()
│
├─ Configure Zaraz → configuration.md
│   ├─ Add GA4/Facebook → tools setup
│   ├─ When tools fire → triggers
│   └─ GDPR consent → consent purposes
│
├─ Integrate with Workers → patterns.md#worker-integration
│   ├─ Enrich context → Context Enrichers
│   └─ Inject tracking → HTML rewriting
│
└─ Debug issues → gotchas.md
    ├─ Events not firing → troubleshooting
    ├─ Consent issues → consent debugging
    └─ Performance → debugging tools
```

## Key Features

- **100+ Pre-built Tools** - GA4, Facebook, Google Ads, TikTok, etc.
- **Zero Client Impact** - Runs at Cloudflare's edge, not browser
- **Privacy Controls** - Consent management, data filtering
- **Custom Tools** - Build Managed Components for proprietary systems
- **Worker Integration** - Enrich context, compute dynamic values
- **Debug Mode** - Real-time event inspection

## Reference

- [Zaraz Docs](https://developers.cloudflare.com/zaraz/)
- [Web API](https://developers.cloudflare.com/zaraz/web-api/)
- [Managed Components](https://developers.cloudflare.com/zaraz/advanced/load-custom-managed-component/)

---

This skill focuses exclusively on Zaraz. For Workers development, see `cloudflare-workers` skill.