This repository's agent instructions live in AGENTS.md, which is the cross-agent standard and the file to edit. This file exists only because Claude Code loads CLAUDE.md and not AGENTS.md, so without it the routing hub never reaches the session.
@AGENTS.md
Nothing about the methodology belongs in this file. A rule written here is invisible to every other assistant working in this repository, and it fails quietly: Claude Code follows it while the others carry on without it, and nothing reports the disagreement. Rules go in AGENTS.md.
Instructions that are genuinely specific to Claude Code — and only those — can be added below the import. This file is yours: clue init never overwrites it and clue migrate never rewrites it.
AGENTS.md routes; it does not restate the method. The rules an agent must follow are in the clue-* skills under .agents/skills/, mirrored to .claude/skills/ so Claude Code lists them. Read the relevant skill before deciding how a change is shaped — the skill is what binds, and a rule that is only in your own /docs is this repository's local bookkeeping rather than something the method requires.
The skills are generated and version-stamped. Do not hand-edit them; clue migrate upgrades them when a new pair is released.