What problem does it solve? AI models often default to outdated framework patterns from training data, shipping deprecated APIs (like Supabase getSession instead of getUser) that require follow-up fix tasks. This Skill verifies that the patterns about to be used for a given framework and version match current official documentation before greenfield code is written. ## Core Features & Use Cases - Official-docs verification: Fetches current recommended patterns, deprecated patterns, and breaking changes per framework version from official docs, release notes, and migration guides. - Project-level caching: Persists verified guidance in CURRENT_PATTERNS.md so subsequent tasks reuse it instead of re-fetching, with a 30-day staleness threshold. - Completeness enforcement: Every requested framework is reported as verified or unverified with a reason, so nothing is silently dropped. - Use Case: When impact-analysis flags greenfield work on a Next.js and Supabase project, run this Skill to confirm current auth and data-fetching patterns, cache them, and unblock implementation planning. ## Quick Start Verify the current official patterns for the frameworks in my project and cache the results in CURRENT_PATTERNS.md.