document-itlisted

# document-it ## What this skill does This skill creates, maintains, and reviews explanatory architecture guides — in Markdown, a Claude Artifact, or both — from already-available, sufficient, and current understanding. It is independently callable: a documentation request does not require a fresh investigation to begin. | User intention | Result | | --------------------------------------------------- | ------------------------------------------------ | | Document existing architecture | A new guide (Markdown, Artifact, or both) | | Reconcile a guide with verified changed reality | An updated existing guide | | Evaluate whether a guide is complete | A findings report, not a rewrite | ## Evidence: reuse first, route to `lab-it` only when needed Before writing or updating anything, establish whether currently available understanding is sufficient and current: - **Sufficient, current evidence already exists** — from this conversation, a recent investigation, or evidence directly verifiable against the real system right now. Use it directly. Don't repeat a full investigation just because a documentation workflow is starting. - **Evidence is missing or stale.** Route the relevant investigation through `lab-it`'s canonical investigation method (its "Shared investigation and decision discipline") rather than dupli