What problem does it solve? Design and implementation work often stalls because engineers lack verified technical facts about vendors, internal systems, and existing code. This Skill collects only the evidence needed for one feature and produces a compact, source-traceable context pack before design begins. ## Core Features & Use Cases - Evidence-Based Context Packs: Gathers official vendor docs first, internal docs second, and code/tests third, recording exact versions, APIs, and constraints. - Layer and Boundary Analysis: Captures layer owners, API contracts, data contracts, and integration seams when a feature spans a vertical slice. - Idempotent Updates: Reuses existing context packs when sources are unchanged and marks affected design blocks stale when sources change. - Use Case: Before designing a new FEAT-12 checkout flow, run this Skill to produce context/FEAT-12.context.md with verified vendor API limits, existing service contracts, and open context questions. ## Quick Start Ask the Agentic SDLC coordinator to gather context for FEAT-12 and produce a context pack before design.