AI agent glossary
AGENTS.md and CLAUDE.md
Plain markdown files in a project that coding agents read automatically to learn its conventions, commands, and rules.
Every time a coding agent starts a session, it knows nothing about your project. A project instructions file fixes that. You write down the things a new teammate would need: how to install and run tests, the folder layout, coding conventions, and things to never do.
AGENTS.md is an open convention read by many coding agents, including OpenAI Codex, Cursor, and others. CLAUDE.md is the file Claude Code reads. Some teams keep one file and point the other at it so every tool gets the same guidance.
Keep these files short and specific. Every line costs context on every session, and vague advice like 'write clean code' changes nothing. Concrete rules, exact commands, and known pitfalls are what pay off.
A CLAUDE.md might say: run the type checker before declaring a task done; database migrations live in db/migrations and must never be edited after merge; use the existing logger instead of console output.
Learn more in the course
Related terms
Go from definitions to doing
Learn to use and build AI agents in 65 practical lessons.