What problem does it solve? When a project ends, most people move on without capturing what they learned, so the same mistakes repeat and hard-won insights are lost. This Skill closes each project by reviewing its full trajectory and distilling lessons that compound into future work. ## Core Features & Use Cases - Trajectory Review: Reads the project's progress.md log end-to-end and reconstructs what actually happened, comparing intended goals against actual outcomes. - Lesson Extraction: Captures only time-saving lessons — tastes, pitfalls, working patterns, and process insights — with honest confidence levels, discarding obvious or redundant findings. - Compounding Storage: Routes lessons to the right places: a reflection note in 40-personal/44-reflections/, profile update suggestions for CLAUDE.md, and wiki accumulation delegated to the wiki-ingest skill. - Use Case: After shipping a workshop program, ask for a retrospective. The Skill reviews the whole project log, identifies that promotion was skipped again (a recurring pattern), writes a reflection note, and proposes a profile update so the next project starts smarter. ## Quick Start Ask the AI to run a retrospective on the finished project and extract the lessons learned.
