# Dispatch and poll

`scripts/relay.mjs` wraps cline's headless JSON mode, captures its NDJSON event stream, and writes
a `result.json`. Run one command, then read one file.

## Before the first run

```bash
command -v cline
cline --version
```

Install `cline`; on macOS/Linux it ships as a native binary, on Windows as an npm package (the
relay launches the `.cmd` shim with `shell:true`). Authenticate with `cline auth`. A headless run
that is not authenticated fails with `status: "failed"` (exit 1).

## Dispatching

```bash
node "<skill-dir>/scripts/relay.mjs" --brief brief.txt --cd /path/to/repo
```

`<skill-dir>` is the installed folder containing this skill's `SKILL.md`.

| Flag | Effect |
| --- | --- |
| `--brief <file>` | Brief path. Omit it to read the brief from stdin. |
| `--cd <dir>` | Working root and child process cwd (default: current directory). |
| `--lane <name>` | Fleet lane from `delegate-setup` config. Applies that lane's dials; fails if the lane's `implementer` is not this relay. Explicit dial flags win. |
| `--provider <name>` | Cline provider name (default: cline's own default). Token-validated. |
| `--model <id>` | Cline model id (default: cline's own default). Provider-local and qualified ids are accepted. Token-validated: letters, digits, `. _ : / -`. |
| `--plan` | Plan mode with `--auto-approve false`; explicit `--auto-approve true` is rejected so Cline cannot auto-approve a switch to act mode. |
| `--auto-approve <bool>` | Cline tool auto-approval. Defaults to `true` in act mode and `false` with `--plan`. |
| `--timeout <dur>` | Relay watchdog (default: `30m`; h/m/s strings). The relay never passes it as Cline's own `-t` / `--timeout`. |
| `--out-dir <dir>` | Artifact directory (default: a fresh directory under the system temp dir). |
| `-h`, `--help` | Print the relay's header help. |

The child cwd pins the workspace; the relay does not pass Cline's `--cwd`. `-v` (verbose) exposes
provider and model in `run_start`. Fresh JSON runs may omit `sessionId`, and the verified headless
JSON path does not support resume, so the relay has no resume flag.

## Artifacts and result fields

Artifacts live outside the repo by default, so they do not appear in `touchedFiles`; an
`--out-dir` inside the worktree can make the artifacts appear there:

- `brief.txt` - the exact brief.
- `events.jsonl` - raw cline stdout events (every event cline emitted).
- `final.txt` - cline's final text (`run_result.text`); absent if none was emitted.
- `stderr.txt` - complete stderr.
- `result.json` - the stable `delegate-relay.result.v1` contract.

`result.json` fields:

- `schema`, `tool` (`"cline"`), `status` (`completed` | `failed` | `timeout` | `aborted` | `cline_unavailable`), `exitCode`, and `signal` (`null` unless the child died on a signal).
- `workdir`, requested `provider`/`model`, `planMode`, `autoApprove`, `clineVersion`, `sessionId`,
  `startedAt`, and `finishedAt`.
- `actualProvider` and the initial `actualModel` from `run_start`; `run_result` can update the model
  and supplies `finishReason`, `usage`, and `durationMs`.
- `briefPath`, nullable `finalPath`, `eventsPath`, and `stderrPath`. `finalPath` is `null` when
  Cline emitted no final text and `final.txt` was not created.
- `sessionId` - parsed from `run_start` when Cline emits one; otherwise `null`. It is observational,
  not a resume promise.
- `finalMessage` - cline's final text (`run_result.text`).
- `touchedFiles` - `git status --porcelain` lines for the **final working tree under `--cd` only**,
  not an attribution of cline's edits: anything already dirty before dispatch shows up too.
  Dispatch from a clean tree when you want the list to read as "what cline changed". `null` means
  git could not report; `[]` means git ran and the tree is clean.
- `stderrTail` - the last 20 non-empty stderr lines on any run that did not complete (`failed`,
  `timeout`, `aborted`), except a launch failure, which reports `failed` with no `stderrTail`.
- `error` - present for launch failures, preflight failures, when the relay watchdog fires
  (`timeout`), and on an `aborted` run.

## Waiting for completion

The relay blocks. Use the orchestrator's background-command facility, or background it in a shell
and poll for `result.json`. The run is done only when the process exits and the file contains a
`status`. The result file is written atomically, so a partial read is impossible.

A pre-run usage error exits 2 and writes no result. A missing `cline` exits 127 and writes
`status: "cline_unavailable"`.

## When a run misbehaves

- **`status: "cline_unavailable"` (exit 127):** cline is not on PATH. Install it, authenticate,
  and re-dispatch.
- **`status: "failed"`:** read `stderrTail`, `stderrPath`, and the tail of `events.jsonl`. Common
  causes: an unknown `--model`, expired credentials, or a provider error. A `run_result` with a
  `finishReason` other than `completed` is failed even if cline exits zero.
- **`status: "failed"` with an `error` mentioning `version preflight`:** the bounded `cline --version`
  probe failed or hung, so cline was never dispatched. Check the install (`cline --version` yourself).
- **`status: "aborted"`:** the relay itself was killed (its parent's timeout, a stopped task, a
  closed terminal) and forwarded the kill to cline. The result is written before the relay exits;
  inspect the working tree before re-dispatching. On native Windows a hard kill of the relay is
  uncatchable (Node supports no `SIGTERM` handler there), so this status may never get written -
  a relay process that is gone without a `result.json` is an aborted run; inspect the working
  tree and `events.jsonl` directly.
- **`status: "failed"` with `signal: "SIGKILL"`:** the host killed the process, commonly through
  the OOM killer or a supervisor timeout. This is not a cline error; check host memory and
  re-dispatch, or split the task into smaller briefs.
- **`status: "timeout"`:** the `--timeout` watchdog killed the run; `error` reads
  `cline did not finish within --timeout <dur>; killed by the relay watchdog`. Increase
  `--timeout` or split the task. On POSIX the relay sends SIGTERM to the process group, waits 10
  seconds, then sends SIGKILL if needed; on Windows there is no escalation phase - the whole
  process tree is felled immediately with `taskkill /pid <pid> /t /f`.
- **Empty `finalMessage`:** inspect `touchedFiles` and the diff. Add a
  `<structured_output_contract>` to the next brief to require a closing report.

## Recovering lost work

`events.jsonl` in the run directory records every event the implementer streamed. If finished
work is lost - the run killed late, or the working tree damaged afterward - read the event log
before re-dispatching: it identifies which files and tool commands were involved, which scopes
what needs redoing. Tool execution events carry the write/edit arguments, but treat any
reconstruction as unverified until it matches a working-tree diff - when the tree still holds the
work, preserve the tree rather than replaying the log.

## What the relay runs

The launch is equivalent to:

```bash
cline --json -v [--provider <name>] [--model <id>] \
  --auto-approve <true|false> [--plan] \
  "Follow the task instructions provided on stdin." < brief.txt
```

Current Cline JSON mode checks for a positional prompt before reading piped input, so the relay
passes the fixed instruction and streams the real brief on stdin. The child process cwd pins the
workspace. Only token-validated provider/model values and fixed text reach the `shell:true` launch
on native Windows; the brief and cwd do not. `-v` (verbose) exposes the requested provider/model.
Before dispatch the relay runs a
bounded `cline --version` preflight (10s cap) so a hung or crashing CLI fails fast and explicitly
instead of hanging the run. In act mode, auto-approval defaults true. With `--plan`, the relay
forces it false because Cline can otherwise auto-approve a switch to act mode. Cline also exposes
sandbox through `--data-dir` / `CLINE_SANDBOX`; the relay leaves that control to the inherited CLI
environment.

## The commit boundary

The relay never commits. Cline edits the working tree; the orchestrator reviews, re-runs the
gates, and commits. See [review-and-land.md](review-and-land.md).

# Multi-task queues

The single-task loop scales to a queue: a removal across layers, a migration across files, or a
refactor sweep. Sequencing and bookkeeping make it trustworthy.

## Run sequentially, one commit per task

Run tasks **one at a time, in dependency order**, landing each after review and gates before
dispatching the next:

```bash
node "<skill-dir>/scripts/relay.mjs" --brief task-01.txt --cd /path/to/repo
```

- Later briefs can rely on earlier work only after it lands.
- One commit per task keeps history reviewable and each step revertible.
- A clean tree before each dispatch keeps `touchedFiles` honest.

Use parallel runs only for genuinely independent tasks in separate working trees. Sequential is
the default because it preserves clean task boundaries.

## Carry decided constraints forward

Fresh cline sessions do not remember earlier tasks. If task 2 chooses a helper name, fixture
location, or interface that task 5 needs, write that fact into task 5's brief.

Cline's verified headless JSON path does not support session resume. Rework and unrelated queue
items both use fresh runs, so every brief must carry the context it needs.

## Keep a progress file

For more than two or three tasks, maintain one progress file beside the work:

- **Status table** - queued / at-implementer / reviewed+committed, with the commit hash.
- **Per-task review notes** - what landed, what you verified, and gate outcomes.
- **Needs your eyes** - design decisions, non-blocking nitpicks, and questions for the human.
- **End-of-run checklist** - the final cross-task verification.

Update it when each task lands, not in one batch at the end.

## Close with a coherence check

After the last task:

- Run the full test/build once more.
- Search repo-wide for the thing the queue changed.
- Replay migrations from a clean state and check drift when applicable.
- Push and open or update the PR only after the final tree is coherent.

## When to stop and ask

Proceed on work that follows from the agreed plan. Stop and surface when:

- A task cannot be completed correctly within its brief.
- Review calls the plan itself into question.
- Gates reveal a problem affecting already-landed tasks.

Report the landed state, commit hashes, and open question, then wait.

# Review and land

The implementer made the changes; you own the judgment. Verify against reality, never the
self-report, and read the diff as generated code because a green gate cannot catch every failure
mode.

## Check tests before trusting gates

If the diff touches existing tests, review those edits first:

- Treat unbriefed test edits as a contract change, not part of the fix.
- Treat newly skipped, disabled, or commented-out tests as failing until proven otherwise.
- Treat loosened assertions the same way: contains/truthy replacing exact matches, broadened
  error types, and widened tolerances all weaken the gate.

## Re-run the gates yourself

`result.json` carries cline's claims, not evidence. Re-run the project's actual test, lint, and
build commands in the working tree and read their output. Passing is necessary, not sufficient.

## Read the diff against the brief

Start with `touchedFiles`, open the diff, and compare it to the brief:

- **Scope creep** - changes the brief excluded.
- **Scope shortfall** - missed behavior, edges, or cleanup.
- **Quiet judgment calls** - defensible but unasked decisions that need review.

## The implementer sweep

Check every diff for patterns gates often miss:

- Hardcoded success or fixture data on a real-work path.
- Catch-all error handling that returns a default instead of propagating or recovering.
- Imports, dependencies, methods, and signatures not present in the installed version.
- Unused imports, uncalled helpers, unreachable branches, and scaffolding comments.
- A second client, error idiom, or logging style beside the repo's existing one.
- Tests that assert internals instead of behavior, or near-duplicate test bodies.
- Optional parameters, config flags, and abstractions with no caller.
- Guards for impossible cases that hide trust-boundary validation.

Send anything blocking back to cline as a delta brief, or fix it in the tree, and report either
choice to the human. Run relevant guard skills if installed.

## The commit boundary

When the gates pass and the diff holds, **the orchestrator commits**, never the implementer.
Write a clear message describing what landed.

From dispatch until that commit, the uncommitted working tree is the authoritative copy of the
implementer's work - the only one you can commit from, and often the only copy at all. Never run
`git checkout`, `reset`, `clean`, or a branch switch in the workspace between those two points -
however messy an interrupted run looks, inspect it first: `git status`, `git diff`,
`git diff --cached` for anything the implementer staged (plain `git diff` is blind to the index),
and commit the intended files explicitly.

## Rework: re-dispatch a corrected brief

Re-dispatch the correction with the needed context:

```bash
echo "The fix is right, but the tests mock the DB session: use the real migrated fixture and
remove the unused import." | node "<skill-dir>/scripts/relay.mjs" --cd /path/to/repo
```

Cline's verified headless JSON path does not support session resume, so each correction is a fresh
run and its brief must restate the required context. Rework gets the same test review, diff review,
and implementer sweep.

## Surface, do not absorb

The human opted into delegation, so committing verified, gate-passing work is the contract. Keep
them in the loop when the work changes shape:

- Report design decisions and defensible-but-unrequested turns.
- Note non-blocking nitpicks you did not block on.
- Stop and ask if correct completion requires going beyond the brief.

For a queue, keep these notes in the progress file described in
[multi-task-queues.md](multi-task-queues.md).

# Writing the brief

A brief is the entire task as cline will see it. It runs in a separate process with **no memory of
your conversation and no shared context** - only the text you send and whatever it can inspect in
the workspace. If a constraint is not in the brief or discoverable in the repo, it does not exist
for cline.

Cline can auto-discover the workspace's `AGENTS.md`. Still restate load-bearing repo constraints in
the brief so the implementer does not have to infer which rules matter for this task.

## Model choice

Cline picks a default model when `--model` is omitted, so a fresh dispatch does not require it.
Pass `--model <id>` only when the human asked for a specific model, or `--provider <name>` to
pick a provider. The relay forwards ids and provider names made of letters, digits, `. _ : / -`
only.

## The shape that works

Use a compact, block-structured brief. State the task, what done means, the few constraints that
matter, and the report cline must return.

```xml
<task>
One or two sentences: the concrete job and where it lives. Then the specifics - current state, what to
change, and explicitly what to leave untouched. The leave-untouched list prevents unrelated refactors.
</task>

<verification_loop>
Run these before finishing and fix anything they surface, do not just report it:
  <the project's real test command>
  <the project's real lint/format command>
  <the project's real build/typecheck command>
Confirm the working tree shows only the intended changes afterward.
</verification_loop>

<action_safety>
Keep changes scoped to the task. No unrelated refactors, renames, or cleanup unless required for
correctness. Do NOT run git add or git commit - the orchestrator commits after reviewing. Leave the
work uncommitted in the working tree.
</action_safety>

<structured_output_contract>
End with a report in this exact shape:
  1. What changed and why
  2. Files touched
  3. Gate outcomes (include test/lint counts)
  4. Anything you deviated on, left open, or want a decision on
</structured_output_contract>
```

Add extra blocks only when the task needs them:

- **Debugging or open-ended fixes** - add `<completeness_contract>` (resolve fully, not just the
  first plausible cause) and `<missing_context_gating>` (find missing repo facts or state what is
  unknown).
- **Research or recommendations** - add `<research_mode>` (separate observed facts, inferences,
  and open questions), and dispatch with `--plan`; the relay pairs it with
  `--auto-approve false` to prevent a switch to act mode.

## Always ask for the report explicitly

The relay builds `finalMessage` from cline's final `run_result` event text. Without a closing
summary, the edits may exist but the result is hard to review. The `<structured_output_contract>`
block makes the expected report explicit.

## Discover the real gates

Read the repo's `AGENTS.md`, `CLAUDE.md`, `Makefile`, `package.json`, or equivalent first and copy
the actual commands into `<verification_loop>`. A brief that says only "run the tests" makes the
implementer guess or skip them.

## Honor repo conventions

Restate the load-bearing house rules in the brief. Cline can inspect the workspace, but the
important constraints should be directly in front of it.

## One task per brief

Keep each brief bounded. One brief -> one cline run -> one reviewed commit keeps the diff and
rollback clean. Split mixed implementation, review, documentation, and roadmap requests into
separate dispatches.

## Premises freeze at dispatch

The implementer starts from the brief's facts and there is no steering channel mid-run. Audit the
fact block before sending - ownership, target branch, constraints, anything a judgment call rests
on. If a premise turns out wrong while the run is live, stop the run and re-dispatch a corrected
brief rather than discounting the output afterward; inspect the working tree and reconcile any
partial or premise-contaminated edits - keep or revert them - before the re-dispatch.

## A worked example

```xml
<task>
In the payments service at services/billing/, the refund path double-charges when a refund is retried
after a network timeout. Make refund submission idempotent: check for an existing refund by idempotency
key before creating a new one. Touch only services/billing/refund.py and its tests. Leave the charge
path, API routes, and data models untouched.
</task>

<verification_loop>
Run and make green before finishing:
  pytest tests/billing/ -q
  ruff check services/billing/
Confirm git status shows only refund.py and its test file changed.
</verification_loop>

<action_safety>
Scope strictly to the refund idempotency fix. No unrelated refactors. Do NOT git add or commit; leave
changes in the working tree for review.
</action_safety>

<structured_output_contract>
Report: (1) the root cause and fix, (2) files touched, (3) pytest and ruff outcomes with counts,
(4) anything left open or needing a decision.
</structured_output_contract>
```

## Brief delivery

The relay streams the full brief on stdin and passes only a fixed positional instruction. Current
Cline JSON mode checks for a positional prompt before reading piped input, so a pure stdin launch
does not work; the fixed instruction satisfies that guard without exposing the brief in argv or
subjecting it to command-line length and Windows shell-quoting limits. Keep secrets out of the
brief anyway: it is preserved in the run's `brief.txt` artifact.

Dispatch with [dispatch-and-poll.md](dispatch-and-poll.md), then review and commit with
[review-and-land.md](review-and-land.md).