How real projects brief Claude Code: the commands, conventions and warnings they put in CLAUDE.md.
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
DAIR-AutoResearchReads credentialsCLAUDE.mdDAIR-AutoResearch/AutoSci/CLAUDE.md
Edit i18n/en/CLAUDE.md, not the active copy at root. Run ./setup.sh --lang en to sync. Full tree: docs/runtime-directory-structure.en.md. Wikilinks: [[slug]]. Slugs are lowercase, hyphen-separated, no spaces. Prefer in order: .venv/bin/python (.venv/Scripts/python.exe on Windows) → active conda env → python3 (python on Windows). Tools auto-load API keys from ~/.env and project-root .env via tools/_env.py.
DAIR-AutoResearchCLAUDE.mdDAIR-AutoResearch/AutoSci/runtime/CLAUDE.md
Single source of truth for what is structurally legal in the wiki. YAML in schema/ and policy/ is the contract; loader.py is the Python access API consumed by tools/lint.py and tools/research_wiki.py. No codegen step — edit YAML, restart the tool. Every change below is YAML-only, zero Python change. Edit, save, done. - New entity type — add a top-level block (<kind>: with dir, fields, optional lifecycle, optional terminal: true) and create templates/<kind>.md.tmpl with the body section skeleton. - New field…
A file Claude Code reads at the start of every session. It holds the commands, conventions and warnings the agent needs for this project.
At the repository root. Claude Code also reads CLAUDE.md files in subdirectories when it works there.
Build and test commands, the project's layout, conventions that aren't obvious from the code, and mistakes to avoid. Short files tend to work better than long ones.
Claude Code reads CLAUDE.md; most other agents read AGENTS.md. Many projects keep one and point the other at it.
