What problem does it solve? When starting a non-trivial task, agents often reinvent solutions or produce shallow link lists instead of grounded research. This Skill teaches the agent to consult a centralized references catalog, select relevant external sources by type and authority, and synthesize findings into actionable patterns with explicit recommendations. ## Core Features & Use Cases - Catalog-driven discovery: Matches the problem against use_when and avoid_when fields in references/*.yaml files, prioritized by source authority (established > vendor > community > curated). - Knowledge-type classification: Distinguishes methodology, heuristic, inspiration, implementation, and discovery sources so each is used appropriately rather than copied blindly. - Structured synthesis output: Produces per-source blocks covering the relevant pattern, why it matters, adaptation guidance, trade-offs, and a recommendation, with confidence levels from CONFIRMED to SPECULATIVE. - Use Case: Before designing an authentication flow, the agent consults the security and UX reference catalogs, extracts principles from OWASP and Laws of UX, and delivers an adapted recommendation with trade-offs instead of a raw link list. ## Quick Start Ask the agent to research how others have solved this problem using the reference catalog and synthesize actionable patterns before starting implementation.
