What problem does it solve? Teams accumulate research findings, guides, and library analyses that are hard to navigate on demand. This Skill answers questions directly from that knowledge base with inline citations, so answers are grounded in documented evidence rather than an AI's general training data. ## Core Features & Use Cases - Teacher mode: Synthesizes findings and guides into a narrative answer with evidence strength, timestamps, and coverage gaps for questions like "what do we know about X". - Builder mode: Produces an ordered recommendation set of guides, patterns, templates, and pitfalls for design questions like "help me design Y". - Honest gap reporting: Flags missing coverage, stale findings, contradictions, and broken references instead of padding with unsourced claims. - Use Case: Ask "what patterns apply to building a multi-agent orchestration layer" and receive a sequenced reading list of guides and templates, each cited by file path, plus pitfalls relevant to your case. ## Quick Start Ask the knowledge base a question such as "what do we know about prompt evaluation strategies" and receive a citation-grounded answer with caveats and gaps.
