What problem does it solve? Code coverage tells you which lines ran during tests, but not whether tests would fail if the code were wrong. This Skill performs static pseudo-mutation analysis — reasoning about hypothetical code changes (boundary flips, boolean inversions, null returns, removed exceptions, arithmetic swaps) and checking whether existing tests would catch each one — to reveal blind spots where tests pass but would keep passing even if the code broke. ## Core Features & Use Cases - Mutation Point Catalog: Systematically identifies boundary, boolean, return-value, exception, arithmetic, and null-check mutation points across .NET, Python, TypeScript/JavaScript, Java, Go, Ruby, Rust, Swift, Kotlin, PowerShell, and C++. - Kill/Survive Classification: Classifies each mutation as Killed, Survived, No coverage, or Equivalent, with per-mutation explanations of why existing assertions miss the change. - Risk-Prioritized Reporting: Produces a mutation score summary plus prioritized recommendations, ranking survived mutations in business-critical logic above low-risk formatting code. - Use Case: After generating tests for a payment calculation module, ask whether the tests would catch an off-by-one in a discount boundary or a removed null guard — the Skill traces each mutation through the test suite and reports exactly which changes would slip through undetected. ## Quick Start Ask the AI to analyze whether the tests for a specific module would catch subtle bugs, for example: "Would my tests catch it if someone changed the boundary condition in OrderCalculator.cs?"
