Blogmark

How to write a great agents.md: Lessons from over 2,500 repositories

via jbranchaud@gmail.com

https://github.blog/ai-and-ml/github-copilot/how-to-write-a-great-agents-md-lessons-from-over-2500-repositories/
LLM AI-assisted Coding

Put commands early: Put relevant executable commands in an early section: npm test, npm run build, pytest -v. Include flags and options, not just tool names. Your agent will reference these often.

I’ve seen big improvements in how agents do with my codebase when I include exactly what command and flags should be used for things like running the test suite.

Cover six core areas: Hitting these areas puts you in the top tier: commands, testing, project structure, code style, git workflow, and boundaries.

Covering these areas is a good starting point. I’m usually not sure where to start and only add things as needed.