What problem does it solve? Code coverage tells you a line executed, not that a test would fail if that line were wrong. This Skill runs mutation testing on a Go file with gremlins — breaking the code one operator at a time — to reveal which changes the test suite would not notice, so you know whether the tests are trustworthy before a refactor. ## Core Features & Use Cases - Automated mutation runs: A script handles module discovery, narrows gremlins to one file, detects build-tag traps that silently produce empty results, and retries timeout-dominated runs automatically. - Action-oriented triage: Survivors are classified as unasserted (needs a sharper assertion), unobservable (semantically equivalent mutant, possibly redundant code), or not worth pinning, with mechanical verification techniques to prove each verdict. - Use Case: Before trusting a pricing module's suite for a migration, run the Skill on the file; it reports that a > → >= boundary mutant survived because no test sits on the threshold value, and tells you exactly which existing subtest to strengthen. ## Quick Start Ask the AI to mutation-test a specific Go file, for example: run mutation testing on internal/lexer/parser.go and tell me whether its tests would catch a defect.