What problem does it solve? It turns completed learning changes into durable, organized Markdown knowledge while preventing uncontrolled writes, duplicate entries, and broken index links in a Learning knowledge base. ## Core Features & Use Cases - Plan Promotion: Analyzes completed changes against existing indexes to produce create, merge, supersede, or archive-only plans staged in promotion-staging/ without writing to the knowledge base. - Apply Promotion: Executes only confirmed plans after drift re-validation, with full rollback snapshots in promotion-rollback/ and reverse-order recovery on validation failure. - Review State Tracking: Marks knowledge as mastered, review_due, or needs_refresh with last-verified and next-review dates for spaced repetition. - Use Case: After finishing a learning session on a new topic, invoke this Skill to merge the validated findings into the domain knowledge file, update the domain INDEX and master INDEX, and schedule the next review date. ## Quick Start Ask the agent to plan a knowledge promotion for the completed change in the Learning context and confirm the exact target paths before applying it.