AGENTS.md Generator
Create AGENTS.md instruction files for Claude Code, Cursor, Codex, and GitHub Copilot.
Loading tool...
What is AGENTS.md Generator?
The AGENTS.md Generator builds instruction files that tell AI coding agents how to work in your repository. Instead of pasting the same rules into every chat session, you commit one AGENTS.md at the project root and agents like Claude Code, Cursor, Codex, and GitHub Copilot pick it up automatically. The tool walks you through seven proven sections - project overview, commands, architecture, code conventions, testing, agent workflow, and safety rules - with sensible defaults and a live Markdown preview, then lets you copy or download the finished file. Everything runs client-side, so your project details stay on your machine.
Key Benefits
- Give every AI coding agent consistent, version-controlled instructions
- Cut onboarding time: agents learn your build, test, and lint commands instantly
- Protect sensitive areas with explicit never-touch safety rules
- Keep conventions enforced in generated code, not just human reviews
- No signup and no server - the file is generated entirely in your browser
Common Use Cases
- •Starting a new repository and want agents productive from the first prompt
- •Standardizing agent behavior across a team or open-source project
- •Onboarding Cursor or Claude Code into an existing codebase with strict conventions
- •Documenting verification commands so agents self-check before declaring done
- •Encoding safety boundaries like protected branches and secret handling
How to Use AGENTS.md Generator
- Describe your project: Enter the project name, primary language, and a one-line description, then pick the AI coding agent you are targeting.
- Choose sections: Toggle the seven standard sections on or off: project overview, commands, architecture, conventions, testing, workflow, and safety rules.
- Fill in the details: Adjust the install, build, test, and lint commands, then add architecture notes, naming and style conventions, testing guidance, workflow rules, and safety constraints.
- Copy or download: Review the live preview, then copy the Markdown to your clipboard or download it as AGENTS.md and commit it to your repository root.
Key Features
- Ready-made section templates: overview, commands, architecture, conventions, testing, workflow, and safety
- Target agent selector for Claude Code, Cursor, Codex, and GitHub Copilot
- Commands table with sensible npm-style defaults you can override
- Live Markdown preview that updates as you type
- One-click copy or download the finished AGENTS.md file
- 100% client-side and offline - no signup, no data leaves your browser
Frequently Asked Questions
What is an AGENTS.md file?
AGENTS.md is a Markdown instruction file placed at the root of a repository that tells AI coding agents how to work with the codebase: which commands to run, which conventions to follow, what the architecture looks like, and which actions are forbidden. Tools like Claude Code, Cursor, Codex, and GitHub Copilot read these repository instruction files automatically when they start working on your project.
How is AGENTS.md different from a README?
A README is written for humans who want to understand and use the project. AGENTS.md is written for AI agents that will modify the code - it focuses on actionable rules: verification commands, code conventions, testing requirements, and safety constraints, rather than marketing or onboarding prose.
Where should I put the AGENTS.md file?
Place it at the root of your repository, next to files like README.md and package.json. Root-level instruction files are picked up automatically by most AI coding agents; some tools also support placing additional AGENTS.md files inside subdirectories for monorepo-scoped rules.
Does this tool send my project details anywhere?
No. The entire generator runs locally in your browser - no accounts, no servers, no tracking of what you type. The generated file is built with client-side JavaScript and only leaves your machine when you copy or download it yourself.