Add an agents.md file to your repo so AI coding assistants understand your dev setup, testing process, and contribution rules.
Build AI tools that read agents.md to automatically onboard themselves to new codebases without asking for manual context.
Document project conventions in a machine-readable format that both humans and AI agents can follow consistently.
| agentsmd/agents.md | elastic/kibana | coleam00/archon | |
|---|---|---|---|
| Stars | 21,034 | 21,076 | 20,925 |
| Language | TypeScript | TypeScript | TypeScript |
| Setup difficulty | easy | hard | moderate |
| Complexity | 1/5 | 3/5 | 4/5 |
| Audience | developer | data | developer |
Figures from each repo's GitHub metadata at analysis time.
AGENTS.md is an open file format standard for giving AI coding agents context and instructions about a software project. Think of it as a README, but written specifically for AI agents rather than human developers, a predictable, dedicated place where an agent can look to understand how to work on the codebase. An AGENTS.md file lives in a project's repository and can contain things like dev environment tips (commands to use, toolchain notes), testing instructions (how to run the test suite, linting requirements), and pull request conventions. The format is simple plain Markdown. The repository also includes a Next.js website at agents.md that explains the format and shows examples.
A standard file format for documenting how AI coding agents should work with your codebase, like a README, but written for AI instead of humans.
Mainly TypeScript. The stack also includes TypeScript, Next.js, Markdown.
Use freely for any purpose including commercial, as long as you keep the copyright notice.
Setup difficulty is rated easy, with roughly 5min to a first successful run.
Mainly developer.
This repo across BitVibe Labs
Verify against the repo before relying on details.