managing-agent-memorylisted

# Managing Agent Memory Reviewer and maintainer agents write notes to `.claude/agent-memory/<agent>/`. Whether those belong in the repository is a **per-project decision** — some projects want them shared, others deliberately do not — and whether the existing ones still deserve to be there is a **judgement call that has to be re-made every time**, because notes rot: the code they describe gets renamed, and the bugs they warn about get fixed. Three failure modes this skill exists to prevent: 1. **Asking the same question forever.** Re-prompting on every commit is noise. 2. **Answering it once for everything.** A decision taken in one project must never silently apply to the next one. There is no global policy. 3. **Committing notes that are now WRONG.** A note claiming an open security hole that has since been closed, or naming a symbol a rename deleted, is worse than no note — it sends the next agent down a path that no longer exists and gets believed because it looks specific. ## STEP 0 — Find the scopes (there is usually more than one) A repository can hold **several independent memory scopes**: one at the repo root and one per sub-project. They are separate stores, written by agents that ran in different working directories, and they must be decided and curated separately. ```bash ROOT=$(git rev-parse --show-toplevel) git -C "$ROOT" ls-files '*.claude/agent-memory/*' | sed 's|\.claude/agent-memory/.*||' | sort -u ``` An empty prefix means the repo root.