session-learnings

Capture session learnings by generating portable artifacts like investigation-log.md and chat-transcript.md that preserve coding decisions and breakthroughs beyond the chat context.

Documenting session decisions is best handled by generating an investigation-log.md that records rationale, dead ends, and breakthroughs alongside a chat-transcript.md for full context.

Generate a Marp deck by running an end-of-session workflow that extracts context and exports deck.md, deck.pptx, and deck.html files summarizing the coding session.

No external dependencies or components are required. The session capture workflow operates independently to extract context, generate artifacts, perform optional cross-model review, and update memory files.

Update MEMORY.md by running an end-of-session workflow that integrates new learnings from the generated investigation log directly into the memory file to feed auto-memory systems.

Use automated session capture at the end of productive AI coding sessions when you need deterministic, searchable documentation artifacts and memory integration rather than relying on manual notes.