memory-merger

To promote lessons learned into permanent instruction files, the memory-merger skill scans LESSONS-LEARNED.md, selects validated entries by frequency and severity, drafts context-aware edits, applies them to target skill files with attribution, and archives the originals.

Automating knowledge management after a retrospective involves parsing accumulated lessons for severity and age, then codifying recurring postmortem learnings directly into team-lead instructions and testing-workflow skills so they are encountered naturally.

Merging lessons into skill files preserves institutional memory by graduating mature, validated insights from a buried lessons file into active workflow instructions, ensuring contributors encounter them naturally during tasks and avoid repeating mistakes.

Yes, you can filter lessons by frequency, severity, age, and category concentration. The skill selects qualifying candidate lessons based on these metrics before drafting contextual edits and applying changes to mapped instruction files.

After lessons are merged into instruction files, the original entries are preserved in an Archived section within the lessons file. The skill simultaneously updates the lessons index to maintain governance and historical traceability.

You should not use an automated lesson merging workflow when lessons lack sufficient frequency, severity, or age metrics to pass quality gates. The skill enforces these quality gates before finalizing any merges to prevent premature codification.