AGENTS.md Generator
Generate AGENTS.md files to guide AI coding agents. Supports Claude Code, Copilot, Cursor, Devin, and more.
AGENTS.md Preview
6 lines# AGENTS.md ## Development Environment - Package manager: `npm`
Compatible AI Coding Agents
AGENTS.md is supported by 60,000+ open-source projects and all major AI coding agents. Place the file in your project root.
How to Use This Tool
- Choose a Quick Preset to auto-fill for popular stacks like React + TypeScript, Next.js, Python + FastAPI, Go, or Rust. Or start from scratch with Custom.
- Fill in the Project Overview: name, description, and tech stack. This helps AI agents understand what the project is about.
- Add Development Environment details: package manager, dev/build/test/lint commands. These are the most critical fields — agents use these commands constantly.
- Describe your Codebase Structure so agents know where to find and place files.
- Define Code Style & Conventions: coding patterns, naming conventions, and error handling rules that agents should follow.
- Add Testing Instructions, PR & Commit Guidelines, and Security Considerations to shape how agents contribute to your project.
- Preview the generated AGENTS.md on the right panel. Click Copy or Download to save the file to your project root.
Why Every AI-Assisted Project Needs an AGENTS.md
AGENTS.md is an open standard for guiding AI coding agents. Think of it as a README for AI — a dedicated, predictable place to provide context and instructions that help AI systems work effectively on your project. It was created by the community and is backed by the Agentic AI Foundation (AAIF) under the Linux Foundation.
Without an AGENTS.md, AI coding agents like Claude Code, GitHub Copilot, Cursor, and Devin have to guess your project's conventions. They may use the wrong package manager, write tests in the wrong framework, or violate your naming conventions. An AGENTS.md eliminates this guesswork by providing structured instructions that all agents understand.
Over 60,000 open-source projects have adopted AGENTS.md since its introduction in 2025. It's supported by every major AI coding tool: Claude Code, GitHub Copilot, Cursor, Windsurf, Devin, Gemini CLI, VS Code AI, and more. The format is simple Markdown — easy to write, easy to maintain, and works everywhere.
Best practices: place AGENTS.md in your project root. For monorepos, you can also place sub-AGENTS.md files in subpackages — agents read the nearest file in the directory tree. Keep it focused on actionable instructions: build commands, test commands, code style rules, and security constraints. Update it as your project evolves.
Last updated: February 2026
FAQ
What is AGENTS.md?
AGENTS.md is an open standard for guiding AI coding agents. It's a Markdown file placed in your project root that tells AI tools like Claude Code, GitHub Copilot, Cursor, and Devin how to work with your codebase — including build commands, code conventions, testing instructions, and project structure.
Which AI coding tools support AGENTS.md?
AGENTS.md is supported by all major AI coding agents: Claude Code (Anthropic), GitHub Copilot (GitHub), Cursor, Windsurf (Codeium), Devin (Cognition), Gemini CLI (Google), Aider, VS Code AI, and more. Over 60,000 open-source projects have adopted it since 2025.
Where should I place the AGENTS.md file?
Place AGENTS.md in the root directory of your project. For monorepos, you can also place additional AGENTS.md files in subpackages. AI agents read the nearest file in the directory tree, with the closest one taking precedence.
How is AGENTS.md different from CLAUDE.md or .cursorrules?
AGENTS.md is a universal, tool-agnostic standard supported by all major AI coding agents. CLAUDE.md is specific to Claude Code, and .cursorrules is specific to Cursor IDE. If you use multiple AI tools, AGENTS.md ensures consistent behavior across all of them. You can still use tool-specific config files alongside AGENTS.md for tool-specific features.
How does AGENTS.md work in monorepos?
In monorepos, you can place AGENTS.md files at multiple levels — project root and inside subpackages. AI agents read the nearest AGENTS.md in the directory tree, with the closest file taking precedence. This lets you define global conventions at the root and package-specific rules in subdirectories.