What problem does it solve? Over time, a knowledge base of engineering learnings (docs/solutions/) drifts out of sync with the codebase: file paths move, recommended fixes become outdated, and duplicate documents silently contradict each other. This Skill audits those documents against the current code and keeps the knowledge base accurate. ## Core Features & Use Cases - Drift Detection and Classification: Cross-references each learning doc against the current codebase and classifies it as Keep, Update, Consolidate, Replace, or Delete based on evidence. - Interactive and Autofix Modes: Runs interactively with one-at-a-time questions for ambiguous cases, or fully unattended via mode:autofix which applies safe actions and marks uncertain docs as stale. - Document-Set Analysis: Detects overlapping, superseded, or contradictory docs and consolidates them into a single canonical source of truth. - Use Case: After months of refactoring, run the refresh to find learnings referencing deleted files, merge two docs covering the same auth problem, and generate a report of applied and recommended actions. ## Quick Start Ask the AI to refresh the experience library by reviewing docs/solutions/ for outdated or duplicate learnings, optionally passing a scope like a module name or mode:autofix.
