What problem does it solve? Before changing code, teams often guess at scope and miss how a request fits the repository's actual architecture, tests, and decisions. This Skill performs a read-only inspection of the current repository and returns a bounded implementation plan grounded in real evidence, preventing mis-scoped or contradictory changes. ## Core Features & Use Cases - Policy-first planning: Reads the project's designated policy or rules documentation before inspecting implementation details, and flags conflicts with active decisions as needs_approval. - Evidence-backed scoping: Searches symbols, exports, callers, tests, and configuration before opening only the function bodies relevant to the request, avoiding duplicate state or logic. - Structured result contract: Returns status, summary, proposedChanges, verification, and risks fields covering affected areas, tests, manual checks, failure modes, and exclusions. - Use Case: Before implementing a new API endpoint, run this Skill to confirm the change fits existing module ownership, identify which tests must be updated, and surface any contradiction with recorded project decisions. ## Quick Start Ask the agent to inspect this repository and produce a read-only, evidence-backed implementation plan for the change you are considering.
