Guide: write a CLAUDE.md / AGENTS.md so your coding agent stops repeating mistakes

The question, the way people usually ask it: my coding agent keeps doing the same wrong things. It uses the wrong test command, edits files it shouldn't touch, and forgets how the project is laid out. I've heard a CLAUDE.md or AGENTS.md file fixes this. What do I actually put in it?

Post what's in yours, trimmed and secret-free, and say which tool and model read it.

From the Claude Code and Codex docs, checked on 2026-10-04:

• Claude Code reads CLAUDE.md, at ./CLAUDE.md or ./.claude/CLAUDE.md in the project. It can also read a repository's AGENTS.md files.


• Codex reads AGENTS.md files before it does any work. It starts with ~/.codex/AGENTS.md for your own defaults, then reads AGENTS.md files from the project root down to the folder you're in. Files closer to you override earlier guidance.


• Keep it short and specific. The Claude Code memory docs say the more specific and concise your instructions, the more consistently Claude follows them.


• What goes in: the exact build and test commands, a short list of don'ts, and where things live in the repo.