What problem does it solve? Small-looking code changes often break things far outside the diff, and a convincing-sounding risk writeup can be wrong. This Skill finds the hidden downstream breakage of a change and proves the one fact its safety depends on by executing real code instead of trusting a narrative. ## Core Features & Use Cases - Hidden Breakage Discovery: Traces what grep and symbol search miss, such as API response shapes, DB columns, wire formats, feature flags, and code several hops downstream. - Evidence-Ranked Verification: Ranks each safety claim on a five-level confidence ladder, from assertion up to reproduction in the running app, and pushes key facts toward executable proof. - Executable Proof Scripts: Writes and runs a small script or test that calls the real shipped code and fails loudly if the safety assumption is wrong. - Use Case: Before merging a PR that changes cache invalidation logic, use this Skill to find the single fact the change is safe because of, prove it with a runnable script, and get a report of confirmed risks, cleared risks, and the cheapest pre-merge 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.
