compound-learnings

To turn session learnings into durable workflow patterns, analyze accumulated session data to identify recurring insights across at least three diverse sessions, then consolidate them into proposed permanent artifacts like rules or skills.

You need to consolidate project learnings into permanent artifacts when you have accumulated multi-session insights and want to scale improvements, requiring at least three learnings across diverse sessions to trigger pattern detection.

You can generate rules or skills from recurring project learnings by applying consolidation logic to your accumulated session data, which proposes structured artifacts for your explicit review and approval before creation.

Yes, you need a minimum of three learnings across diverse sessions within a multi-session project before the analysis mechanism can identify recurring patterns and propose durable artifacts.

Yes, artifact generation requires explicit user approval before creating any rules, skills, or hooks, ensuring you review proposed patterns derived from the learnings cache before they become permanent.

The best way to manage a learnings pipeline for multi-session projects is to use project-scoped consolidation logic that tracks, generates, and archives artifacts from your learnings cache folder after detecting recurring patterns.