curate

> **Manual-trigger skill.** `disable-model-invocation: true` keeps the model from > running docs audits automatically mid-task. When installed from this plugin, > invoke it deliberately as `/agent-docs:curate` when you want to clean up the > docs/ tree. Do not add hooks, background tasks, auto-trigger behavior, runtime > storage, vector databases, MCP integration, or external memory systems. Review the `docs/` knowledge base and produce a structured `Docs Health Report` using the workflow below. ## What this does This is the complement to `/agent-docs:remember`: - `/agent-docs:learn` proposes verified knowledge additions to `AGENTS.md` or `docs/`. - `/agent-docs:remember` audits `AGENTS.md` memory surfaces (prompt-resident). - `/agent-docs:curate` audits the `docs/` knowledge base (pull-based). The economics differ and so do the rubrics. `AGENTS.md` pays per prompt line, so `remember` uses a high value threshold. `docs/` is read on demand and may legitimately be longer or more detailed, so `curate` judges **navigability, link integrity, structure, and drift** — not brevity for its own sake. Derivability alone never makes useful documentation invalid. Read and apply the shared [Knowledge Admission Policy](../../references/knowledge-admission.md) and [Documentation Structure Reference](../../references/doc-structure.md). Do not invoke `learn`; all knowledge workflows independently use the shared policy. **Scope guard — `docs/` only, never `AGENTS.md`.** Audit files und...