corgispec-memory-extract

To extract reusable patterns from completed AI coding sessions, you distill non-trivial approaches from finished changes into long-term wiki memory. This creates reusable pattern pages under the wiki/patterns directory for future reference.

Writing a session summary for a completed change involves capturing the timeline, key decisions, pitfalls, and outcomes into a historical traceability file. This ensures the wiki sessions record is preserved without overwriting existing summaries.

Cross-session context lifecycle bookkeeping updates wiki index pages and moves the change from Active to Recently Shipped. It resets the session-bridge memory for the archived change while keeping other active change entries intact.

You do not need full archive closure to perform knowledge distillation from a finished task. Pattern extraction and session summary generation can be applied when knowledge extraction is needed independently of the final archiving step.

You can clear individual archived entries in the session bridge without affecting active changes. The knowledge extraction process selectively clears only the archived change entries in memory, leaving all other ongoing work intact.