LLM Mart Basic
@llm-mart · Joined Jun 2026
Apply better-data's security discipline when touching Secret, EncryptionEngine, #[Sensitive], #[Encrypted], MetaKeyRegistry::register, RequestSource guards, or user_pass handling. Loud-over-silent — missing key throws, tampered ciphertext throws, unknown strict-whitelist field th
Add a new sink to better-data — code that writes
Add a new source adapter to better-data — code that reads from a WordPress data store the library doesn't cover yet (comments, attachments, transients, custom tables). Mirror the canonical shape PostSource / UserSource / TermSource use — a non-final class with static hydrate(int|
Add a new validation rule to better-data — implement
Configure better-route 1.1 AtomicIdempotencyMiddleware for side-effectful POST, PUT, PATCH, or DELETE routes where concurrent duplicate execution must be prevented. Use for WpdbAtomicIdempotencyStore, lease-aware reservations, reservation_token schema migration, Idempotency-Key v
Configure Better Route 1.1 audit events and safe enrichment. Use when logging route outcomes, authenticated identity, hashed idempotency keys, trusted client IPs, domain action metadata, or ensuring telemetry failures cannot change API behavior.
Configure Better Route 1.1 authentication with JWT, custom bearer tokens, WordPress Application Passwords, or cookie nonces. Use when protecting routes, mapping verified claims to WordPress users, enforcing scopes, or consuming the shared AuthContext identity.
Configure better-route 1.1 CORS for browser, mobile, and embedded WordPress REST clients. Use for CorsPolicy, CorsMiddleware, WordPressCorsBridge, allowed origins/methods/headers, credentials, OPTIONS preflight, Authorization, X-WP-Nonce, Idempotency-Key, If-Match, If-None-Match,
Use Better Route 1.1 cryptographic helpers for secure random tokens, Hex/Base64/Base64URL encoding, strict Base64URL decoding, and constant-time secret comparison. Use when implementing nonces, state, PKCE, opaque tokens, or signature comparisons.
Produce and consume better-route 1.1 structured errors. Use for ApiException, Response, ErrorNormalizer, ResponseNormalizer, WP_Error conversion, OAuth RFC 6749 error_format, status/code/details/headers, Retry-After, validation_failed, idempotency/rate/optimistic-lock/Woo errors,
Add better-route 1.1 ETag and If-None-Match handling to GET or HEAD routes. Use for ETagMiddleware, strong or weak validators, custom etagResolver, WP_REST_Response preservation, comma-separated validators, wildcard matching, 304 responses, Cache-Control, proxy-stripped ETag trou
Configure Better Route 1.1 HMAC authentication for webhooks and server-to-server REST requests. Use when signing request timestamps, methods, paths, raw bodies, optional query strings, rotating key IDs, or consuming the shared HMAC AuthContext identity.
Configure better-route 1.1 replay-cache idempotency with IdempotencyMiddleware and Idempotency-Key. Use for ArrayIdempotencyStore, TransientIdempotencyStore, WpdbIdempotencyStore, installSchema, identity-aware canonical keys, body fingerprints, key conflicts, replay headers, key
Install better-route from Packagist or migrate a WordPress integration to better-route 1.1. Use when adding better-route/better-route, changing the Composer constraint to ^1.1, upgrading from 1.0 or pre-1.0 releases, diagnosing new 403 route responses, migrating atomic idempotenc
Configure Better Route 1.1 RS256 or ES256 JWT verification from a local or HTTPS JWKS. Use when integrating OIDC/OAuth bearer tokens, selecting keys by kid, validating issuer/audience/lifetime, or operating JWKS caching and refresh behavior safely.
Configure Better Route 1.1 trusted-proxy client IP resolution and CIDR allowlists. Use behind Cloudflare, nginx, load balancers, or reverse proxies when authorization, rate limiting, or audit data depends on the real client IP.
Generate or serve better-route 1.1 OpenAPI 3.1 documents from Router/Resource/Woo contracts. Use for OpenApiExporter, OpenApiRouteRegistrar, contracts, contractsFromSources, route args to parameters, explicit parameter overrides, custom responses, OPTIONS 204, strictSchemas, comp
Configure Better Route 1.1 optimistic locking for REST writes with If-Match or version parameters and an atomic per-resource critical section. Use when preventing stale updates, lost writes, or two cooperating Better Route requests from passing the same version check concurrently
Add Better Route 1.1 ownership authorization to raw routes and Resource DSL endpoints. Use when authenticated users may access only their own records, orders, profiles, memberships, tokens, or subscriptions.
Configure better-route 1.1 RateLimitMiddleware with atomic fixed-window storage. Use for WpObjectCacheRateLimiter, TransientRateLimiter, persistent external object cache checks, wp_cache_incr, MySQL named locks, identity/native WordPress/IP keys, trusted proxies, Retry-After and
Fourteen posts of being wrong in production, compressed to checkboxes
Healthy nodes, a quiet network, 300 restarts in three days, and a latency budget measured in milliseconds
Discovery worked. Ping worked. Every TCP connection timed out, and later the tunnel only worked when someone had a terminal open.
Every VM came back. The cluster did not. Declarative systems converge on config, and the datapath isn't config.
A surprising share of AI-in-the-terminal failures aren't the AI. They're zsh, and a version of bash from 2006.
A Claude Code plugin turns standalone project configuration into a namespaced, installable extension that teams and communities can update as one unit.
None of the safety came from the model. It came from six boring habits.
Skills package instructions and references. Subagents run work in a separate context and return results. They solve different problems and can be composed deliberately.
Six hours in, one step left, everything green, and the incident that didn't happen
CLAUDE.md carries persistent project context. Skills load reusable procedures when relevant. Separating stable facts from task-specific workflows keeps both easier to maintain.
Twenty minutes recovering secrets that never existed, and the one sentence from a human that ended it
An API request routing a model's tool call through an approval gate to a remote MCP server
31 config keys, two audits, and why the first one was wrong in both directions
The official MCP Registry stores standardized server metadata rather than package code. Publishers verify a namespace, describe installation or remote access, and submit immutable versions.
Everyone looks at the Dockerfile. The file that actually leaked the key was the project file.
Remote MCP authorization uses established OAuth standards, but secure integration still requires issuer validation, least-privilege scopes, protected token handling, and server-side enforcement.
"Copy it over and switch the reference" is two steps, and the outage lives in the one nobody checks
stdio fits local processes and prototypes. Streamable HTTP fits hosted services and shared integrations. The right choice follows where the capability runs and who must reach it.
The most important rule wasn't about what I could change. It was about what I was allowed to display.
Tools perform operations, resources expose readable context, and prompts provide reusable templates. Choosing the correct primitive makes an MCP server easier to understand and govern.
/install
install
Install a rule from the catalog
/list
list
List all rules with their status
/catchup
catchup
Context recovery after session breaks. Summarize recent git changes, extract key decisions, and identify pending work.
/justify
justify
Audit changes for additive bias, Iron Law compliance, and minimal intervention. Justifies every change against a "simplest safe change" standard.
/karpathy-check
karpathy-check
Pre-flight gate walking four Karpathy principles (think first, simplicity, surgical edits, goal-driven). Use before implementation.
/stewardship-health
stewardship-health
Display stewardship health dimensions for one or all plugins
/structured-review
structured-review
Start a structured review workflow with evidence logging and formatted output
/reinstall-all-plugins
reinstall-all-plugins
Uninstall and reinstall all Claude Code plugins to refresh cache and resolve version mismatches.
/update-all-plugins
update-all-plugins
Update all installed Claude Code plugins from all marketplaces.
/verify-plugin
verify-plugin
Verify plugin behavioral contract history via GitHub Attestations (SLSA)
/garden
garden
Manage and analyze digital gardens for evolving knowledge bases, including health metrics.
/navigate
navigate
Search and navigate across memory palaces to find and retrieve stored information.
/palace
palace
Create and manage memory palaces, the unified storage for knowledge captured from research sessions.
/review-room
review-room
Capture and retrieve PR review knowledge in project memory palaces.
/skill-logs
skill-logs
View and manage skill execution memories stored by memory-palace.
/close-issue
close-issue
Analyze if issues (GitHub/GitLab) can be closed based on commits
/create-issue
create-issue
Create a new GitHub issue with formatted description, labels, and optional reference links
/update-labels
update-labels
Reorganize GitHub issue labels into type, priority, and effort groups.
/oracle-setup
Oracle setup
Provision the oracle ML inference daemon
/analyze-tests
analyze-tests
Analyze Python test suites for quality, coverage, and improvement opportunities.
Your car as a chat-room agent: Raspberry Pi 5 + dashcam + local AI. CodeWatch's sibling for the garage.
23 views 0 likesThe enterprise gateway for autonomous agents. Identity management, per-channel isolation, credential vault, per-session tamper-evident audit log.
13 views 0 likesAI agent for Home Assistant — talk to your home, create automations in plain language, analyze cameras with face recognition, and get proactive anomaly alerts.…
14 views 0 likesA desktop app to prototype agent ideas, inspect every harness step, replay failures, and evaluate performance, all in one place. Local-first, cloud-ready for ma…
14 views 0 likesLocal-first AI memory you can see, edit, and override — portable across Claude Code, Codex, Cursor, Windsurf, and other MCP coding tools.
14 views 0 likesWeb & desktop (Electron) client for the OCTO open workplace — one React + TypeScript codebase shipping browser and PC surfaces, with first-class AI agent UX.
13 views 0 likes同花顺官方 A股金融数据服务,提供股票实时行情、历史行情、财务报表、指数、板块、涨停等数据,适用于 AI Agent、量化研究和应用开发,支持 API、MCP、CLI 和 Python。Official Tonghuashun (HiThink) A-share financial data service provi…
19 views 0 likesOpen-source desktop SQL workspace for PostgreSQL, MySQL/MariaDB, SQLite and 15+ more databases like DuckDB, ClickHouse, Redis and Firestore. Built-in MCP server…
12 views 0 likes🤖 Local-assist BOSS Zhipin CLI for AI agents — search, welfare filtering, shortlist, JSON-envelope output; low-risk & compliant by default.
15 views 0 likesThe local-first sidebar AI agent for ComfyUI — runs on your own Claude OR ChatGPT subscription (no API keys, no extra LLM costs). Drives your live graph: edits,…
13 views 0 likesPersistent memory for Claude Code — identity, context, and continuity across sessions
16 views 0 likes🔬 A curated collection of 23,000+ agent skills for empirical research across 8 social science disciplines. | 精选 23,000+ AI Agent 技能库,覆盖8大社会科学学科的实证研究。CoPaper.AI…
17 views 0 likesA curated list of plugins, skills, MCP servers, patch/profile layers, orchestrators & UIs for DeepSeek Harness (DSH). Visualization · PPT · Coding · Agents · Lo…
25 views 0 likesWorld Memory Protocol.
23 views 0 likesRun your own organization of agents.
26 views 0 likesThe most RAM efficient harness
26 views 0 likesAI-native, free, open-source alternative to Jira, Trello, ClickUp & Monday. Built for Scrum teams where humans and AI agents collaborate as equals — on the same…
13 views 0 likesFree open-source desktop app that scrapes JAV metadata and generates NFO + cover art for Jellyfin, Emby & Kodi. No Docker, no CLI — one-click install on Windows…
15 views 0 likesA cross-platform desktop app to manage Agent Skills in one place and sync them to multiple AI coding tools’ global skills directories — “Install once, sync ever…
14 views 0 likes天枢 (Tianshu) 是一个基于harness工程的终端编程智能体运行时(Tui X Gui),针对DeepSeek V4 做了前缀缓存工程优化(长会话实测稳态命中率 97–99%)和深度适配。它跳出了传统 AI 编程助手把大模型仅当成“工具”的局限,基于认知虚拟机 (CVM)、自感知层和信息素(Stigmergy…
11 views 0 likes