Run assertion-driven mutation testing on business-rule critical paths and generate a QA mutation report.
npx skills add https://github.com/mgeer/qa-skills --skill qa-mutate
Please help me install this Agent Skill. Skill: qa-mutate Source: https://github.com/mgeer/qa-skills/tree/main/qa-mutate Command: npx skills add https://github.com/mgeer/qa-skills --skill qa-mutate
Perform assertion-driven mutation testing to evaluate test suite effectiveness. Use this skill when the user wants to assess test quality via mutation testing, find test blind spots, get a confidence score for their test suite, or verify that tests truly catch business rule violations. Triggers on phrases like "mutation testing", "qa mutate", "变异测试", "test quality score", "测试信心评分", "test effectiveness".
Invoke the qa-mutate workflow to run assertion-driven mutations and generate a mutation report.
High-intent search queries and answers about installing and using this skill.
Assertion-driven mutation testing works by mutating business-rule critical paths and checking if tests detect the changes. Surviving mutations highlight test blind spots where assertions fail to catch business rule violations.
Run scoped mutation testing by targeting a specific module (e.g., qa-mutate billing). Scoped runs apply a maximum of 10 mutations to the target module and restore files via git after each evaluation.
Mutation testing requires a clean baseline: an assertions list from _qa/qa-analysis.md or _qa/assertions.yaml, a codebase and tests with a defined mutation plan, and git available to restore files after each mutation.
Yes, you can perform a full mutation test run across all modules with a maximum of 20 mutations. This full run evaluates test suite effectiveness globally while using git to restore files after each mutation.
Scoped mutation testing targets a single module with up to 10 mutations, while full runs cover all modules with up to 20 mutations. Both evaluate test effectiveness and restore files via git after each mutation.
mgeerTranslate project requirements and code behavior into discrete, verifiable assertions and generate a Markdown QA analysis report in _qa/.
JamkrisRun sandbox-mode regression tests to prevent AI-induced code regressions.
mazreanDesign test cases using boundary value analysis and equivalence partitioning.
rudironsoniAnalyze .NET test quality using Coverlet, ReportGenerator, and Stryker.NET.
