story-done

Story acceptance criteria verification checks each story checkbox via file existence, tests, and grep-based hardcoded value detection. It validates implementation coverage against requirements and prompts manual confirmations when automated checks are insufficient.

Design deviation detection flags GDD and ADR drift by comparing current implementation against the latest registry requirements and related decision constraints. It warns developers when implementation diverges from documented architecture intent.

End-of-story gatekeeping applies to Logic, Integration, Visual/Feel, UI, and Config/Data work during sprint production. It requires reading the target story, resolving review mode, and performing evidence checks before updating story state.

Yes, you can optionally spawn QA and code review gates during the story completion process. These gates run alongside evidence checks and deviation checks to ensure implementation quality before the story status updates to Complete.

Manifest staleness checks warn developers when configuration and data manifests lag behind current implementation. Checking manifest staleness prevents marking stories complete when required data or config updates are missing from the build.

Lean review mode quickly validates criteria coverage and surfaces what comes next from the sprint plan. It provides a fast path for verifying story acceptance criteria without spawning optional QA and code review gates.