Featured
Launch your own OpenClaw agent in one click
Sponsored placement
SetupClaw: done-for-you OpenClaw for founders & exec teams
Sponsored
One API to scrape, enrich, and extract the internet.
Sponsored placement
CLN.Work — Stop prompting, start hiring AI employees
Limited-time offer
Launch OpenClaw or Hermes on Hostinger in 60 seconds
Featured
Launch your OpenClaw wrapper and start making money today
Free tool
AGENTS.md Generator
Give Codex, Cursor, and every other AI coding agent the context they need. Fill in your project, setup commands, code style, and PR rules, then copy or download a clean AGENTS.md.
Project details
AGENTS.md preview
# Acme Web App
A Next.js dashboard for the Acme analytics product. Server components by default, client components only when interactivity is required.
## Project overview
- Stack: Next.js 14, TypeScript, Tailwind CSS, Postgres, Prisma
## Setup commands
```bash
pnpm install
cp .env.example .env.local
pnpm db:migrate
```
## Build and test
- Build: `pnpm build`
- Test: `pnpm test`
- Lint: `pnpm lint`
Run the full test suite before finishing a task and fix any failures you introduced.
## Code style
- TypeScript strict mode
- no any
- Prefer function components and hooks
- Single quotes
- no semicolons (Prettier enforced)
- Colocate tests next to source files
## PR instructions
- Title format: [scope] short description
- Run lint and the full test suite before committing
- Keep PRs focused on a single change
## Security considerations
- Never read or log values from .env files
- All user input must go through the validators in lib/validation
## Never do
- Never commit secrets or .env files
- Never edit generated files under prisma/generated
- Never add a dependency without checking for an existing one
What AGENTS.md is and how this generator works
AGENTS.md is an open format for guiding AI coding agents, often described as a README for agents. It is plain Markdown placed at your repository root, and it gives agents one predictable place to find the context a human would pick up from onboarding: how to set the project up, how to build and test it, what the code style is, and what the rules are for commits and pull requests. Tens of thousands of open-source projects already ship one, and it is read by OpenAI Codex, Cursor, GitHub Copilot, Google Jules, Gemini CLI, Aider, Zed, Devin, and many other tools. Our Codex CLI guide shows the file in action.
This generator turns a short form into a sectioned file that follows the conventions on agents.md: a project overview, setup commands in a runnable code block, build and test commands, code style rules, PR instructions, and optional security considerations, plus an explicit list of things agents should never do. Type into any field and the preview updates live; copy the result or download a ready-to-commit AGENTS.md. The standard has no required fields, so empty sections are simply omitted.
AGENTS.md and CLAUDE.md solve the same problem for different tools. Claude Code reads CLAUDE.md as its project memory, while the wider agent ecosystem standardized on AGENTS.md. Many teams keep both, or keep one file and symlink the other to it, since the content overlaps almost entirely. If you use Claude Code, pair this tool with our CLAUDE.md generator and our complete Claude Code guide.
Frequently asked questions
What is an AGENTS.md file?
It is a plain Markdown file at your repository root that AI coding agents read automatically for project context: setup commands, build and test instructions, code style, PR conventions, and hard rules. It is an open standard with no required fields, so any headings work, but predictable sections make it easier for both agents and humans.
Which tools read AGENTS.md?
A long and growing list: OpenAI Codex, Cursor, GitHub Copilot, Google Jules, Gemini CLI, Aider, goose, opencode, Zed, Warp, VS Code, Devin, JetBrains Junie, RooCode, and more. That breadth is the main reason to keep one: a single file covers every agent your team or contributors might use.
AGENTS.md vs CLAUDE.md: which do I need?
CLAUDE.md is Claude Code's native memory file; AGENTS.md is the cross-tool standard used by Codex and most other agents. The content is nearly identical, so many repositories keep one file and symlink the other to it. If your team only uses Claude Code, start with CLAUDE.md; if contributors use mixed tools, ship AGENTS.md too.
Can I use multiple AGENTS.md files in a monorepo?
Yes. Place one at the repo root and add more inside packages that need their own rules. By convention, the closest AGENTS.md to the file being edited wins, and explicit instructions in the chat prompt override everything in the file.
Does this tool send my data anywhere?
No. Everything runs in your browser. The generator builds the markdown locally, and nothing you type is sent to a server, so you can safely draft internal project details before copying or downloading the file.