What problem does it solve? After an epic ships across many coding sessions, no single session saw the whole change, so cross-story defects, god-class growth, and spec drift go unnoticed. This Skill runs an evidence-based retrospective that reads the epic spec, story files, full diff, commits, and sprint status, then judges the epic against its own acceptance criteria. ## Core Features & Use Cases - Evidence-gathered findings: Every reported finding carries a source reference (file, line, commit, or log); unsourced claims are dropped. - Aggregate diff analysis: Derives architecture deltas, duplication maps, god-class growth, pattern divergence, and spec-to-implementation reconciliation across the full epic range using deterministic scripts. - Acceptance verdict with sprint-status integration: Renders accepted, accepted-with-open-items, or rejected, writes a retrospective document, and surgically updates sprint-status.yaml with action items via a comment-preserving, rollback-safe script. - Use Case: After an unattended multi-story epic run finishes, ask for a retrospective to get a sourced findings report, a machine verdict forced to rejected when stories remain unfinished, and tracked action items for the next sprint. ## Quick Start Ask the assistant to run a retrospective on the most recently completed epic and produce a verdict with sourced findings.
