What problem does it solve? Code reviews often miss breakage that grep and symbol searches cannot reveal, such as changes to wire formats, DB columns, library behavior, or downstream consumers. This Skill finds the hidden blast radius of a change and proves the one fact its safety depends on by executing real code instead of trusting a convincing writeup. ## Core Features & Use Cases - Hidden Breakage Detection: Traces risks that symbol search misses, including API JSON shapes, pinned library versions, microtask timing, feature flags, and code three hops downstream. - Proof by Execution: Identifies the single fact the change's safety depends on and verifies it with a runnable script or test that fails loudly if wrong, rating confidence on a five-step scale from assertion to live reproduction. - Honest Risk Reporting: Separates confirmed risks with real file:line citations from cleared cases, and marks anything unproven rather than overstating it. - Use Case: Before merging a diff that touches a shared cache-invalidation call, use this Skill to find the one safety fact, write a small script that calls the real library function, and hand back a report of proven facts, real risks, and the cheapest regression test. ## Quick Start Ask the AI to analyze the blast radius of your current diff and prove the one fact it is safe because of by running real code.
