AGENTS.md examples from real projects

How real projects brief Codex, Cursor and every other agent that reads AGENTS.md.

Best matches · from page 22Worked for most · soon

VercelAGENTS.md

vgpu / apps

vercel-labs/vgpu/apps/docs/agents.md

/react`. Package-owned Ask AI uses AI SDK v6; migrate local AI SDK code separately from Geistdocs route adapters. - Import source-config helpers from `@vercel/geistdocs/source-config` in `source.config.ts`. Do not import

2.3k7d agoReads credentialsDiscuss

Microsoft AzureAGENTS.md

azure-sdk-for-js

Azure/azure-sdk-for-js/AGENTS.md

Troubleshoot a CI / pipeline failure | `.github/skills/azsdk-common-pipeline-troubleshooting/SKILL.md` | | Create a new package-specific skill | `.github/skills/create-package-skill/SKILL.md` | | Copilot code review (CCR) | `.github/skills/code-review/SKILL.md` (routes to canonical reviewer guidance) | | Review criteria (architecture, perf, security, deps, tests

2.3k19mo agoDiscuss

codejunkie99AGENTS.md

agentic-stack

codejunkie99/agentic-stack/.agent/AGENTS.md

Agent Infrastructure This folder is the portable brain. Any harness (Claude Code, Cursor, Windsurf, OpenCode, OpenClaw, Copilot CLI, Gemini, Hermes, Pi, Codex, standalone Python, Antigravity) can mount

2.3k25d agoDiscuss

Sahir619AGENTS.md

fable-method

Sahir619/fable-method/AGENTS.md

work: the opening paragraph must be readable by someone who never saw the code or the data. Define jargon at first use and translate numbers into meaning ("about twice

2.3k3mo agoDiscuss

numman-aliAGENTS.md

cc-mirror

numman-ali/cc-mirror/AGENTS.md

OPUS_MODEL` - `ANTHROPIC_DEFAULT_HAIKU_MODEL` - Optional: `ANTHROPIC_SMALL_FAST_MODEL`, `ANTHROPIC_MODEL`, `CLAUDE_CODE_SUBAGENT_MODEL` ## Provider Blocked Tools Providers block tools via `settings.json` `permissions.deny` entries (written during variant

2.3k4mo agoReads credentialsDiscuss

jwadowAGENTS.md

kiro-gateway

jwadow/kiro-gateway/AGENTS.md

break compatibility - Clients can ignore added fields - original upstream data is always preserved ### Code Review Reality Check **Your code quality reveals your approach.** Contributions missing tests, using non-English identifiers

2.3k8mo agoReads credentialsDiscuss

777geniusAGENTS.md

agent-teams-ai

777genius/agent-teams-ai/AGENTS.md

src/features/CLAUDE.md](src/features/CLAUDE.md) ## Review guidelines - Treat regressions in agent team messaging, task lifecycle, session parsing, code review UI, and provider/runtime detection as high priority. - For team launch hangs, OpenCode `registered`/`bootstrap

2.2k4mo agoDiscuss

DicklesworthstoneAGENTS.md

mcp_agent_mail

Dicklesworthstone/mcp_agent_mail/AGENTS.md

extracted code snippets. --- ## cass — Cross-Agent Session Search `cass` indexes prior agent conversations (Claude Code, Codex, Cursor, Gemini, ChatGPT, etc.) so we can reuse solved problems. **Rules:** Never run bare

2.2k4mo agoPipes a download into a shellDiscuss

Kritt-aiAGENTS.md

open-kritt

Kritt-ai/open-kritt/AGENTS.md

ESLint/Prettier for JS, Ruff for Python). Match the existing style; don't reformat unrelated code. - **Never commit secrets or local data:** credentials, tokens, `.env`, provider login data, or scan outputs

2.2k2mo agoReads credentialsDiscuss

majidmanzarpourAGENTS.md

threejs-game-skills

majidmanzarpour/threejs-game-skills/AGENTS.md

have to name every specialist skill. The same nine skills serve Codex and Claude Code. The user's requested scope, art style, constraints, and prior decisions take precedence over skill

2.2k3d agoDiscuss

cjo4m06AGENTS.md

mcp-shrimp-task-manager

cjo4m06/mcp-shrimp-task-manager/docs/agents.md

Engineer - Testing → QA Engineer 2. **Consider model strengths**: - Complex reasoning → Advanced models - Code generation → Specialized coding models - Documentation → Language-focused models 3. **Use templates**: - Start with pre-configured templates - Customize based

2.2k18mo agoDiscuss

CloudflareAGENTS.md

templates

cloudflare/templates/AGENTS.md

endpoints** - Each endpoint returns expected status/data 5. **HTML/UI** - Key elements are present in responses --- ## Code Review Patterns ### Logic Bug Detection When reviewing template code, watch for these common issues: **Request

2.1k9mo agoDiscuss

amElnagdyAGENTS.md

delegate-skills

amElnagdy/delegate-skills/AGENTS.md

skills** — skills that let an orchestrating agent drive a separate CLI coding agent as an implementer, then review and land the result. Seventeen implementer skills ship today: `claude-delegate` (Claude

2.1k2mo agoDiscuss

alpic-aiAGENTS.md

skybridge

alpic-ai/skybridge/AGENTS.md

pushing a branch needs no special approval. Only a published GitHub release publishes `latest`. ## Code rules **Biome** handles lint + format (see `biome.json`): - Double quotes, 2-space indent, auto-sorted imports

2.1k28d agoDiscuss

LumiwealthAGENTS.md

lumibot

Lumiwealth/lumibot/AGENTS.md

links. - Keep generated illustration code and invented trading results out of instructional examples. Runnable code and actual result evidence belong in the page itself. - Every generated image must be visually

2.1k19d agoReads credentialsDiscuss

rynfarAGENTS.md

meridian

rynfar/meridian/AGENTS.md

Meridian Meridian bridges supported coding clients to Claude through the Agent SDK and to Antigravity through the official agy CLI. ## Working scope Continue authorized implementation through relevant verification, corrections

2.1k11d agoDiscuss

rynfarAGENTS.md

meridian

rynfar/meridian/docs/agents.md

automatically. Meridian parses OMO's Task tool descriptions to extract subagent names (explore, code-review, etc.) and builds SDK AgentDefinitions so Claude can route to the correct agent. Internal orchestration

2.1k11d agoReads credentialsDiscuss

omnimind-aiAGENTS.md

OmniBot

omnimind-ai/OmniBot/AGENTS.md

prepare publishable code changes. - Codex-created issue fixes should use a bot branch and draft PR targeting the default branch, usually `main`; branch protection and maintainer review control the merge

2k6mo agoReads credentialsDiscuss

yyjeqhcAGENTS.md

webcodex

yyjeqhc/webcodex/AGENTS.md

Implement and review - User outcomes, scope, prohibitions, and acceptance conditions are hard constraints. Suggested files and steps are guidance unless explicitly mandatory; adapt them to verified code and conventions. - Follow

2k3d agoDiscuss

doobidooAGENTS.md

mcp-memory-service / directives

doobidoo/mcp-memory-service/.claude/directives/agents.md

Full automated review (5 iterations, safe fixes enabled) - Quality gate checks before review - Test generation for new code - Breaking change detection **Usage:** ```bash # Full automated review bash scripts/pr/auto_review.sh # Quality gate

2k5d agoReads credentialsDiscuss

CLAUDE.md vs AGENTS.md

About AGENTS.md

What is AGENTS.md?

An open format for instructions to coding agents, read by Codex, Cursor and others. Think of it as a README written for agents.

Where does it go?

At the repository root, with more specific files in subdirectories. Agents read the one closest to the file they're editing.

What should it contain?

Setup and test commands, code style, and the rules a new contributor would need to know.

Does Claude Code read it?

Claude Code reads CLAUDE.md. A one-line CLAUDE.md that points at AGENTS.md covers both.