test-gap-analysis

What it does

Analyze production code in any supported language by reasoning about hypothetical mutations, then confirm them against the real test suite to reveal blind spots where tests pass but would fail if code were broken. It uses language extensions to map concepts like null checks and error handling to the target language and identifies mutation points across boundaries, boolean logic, return values, exceptions, arithmetic, and null checks. It intentionally avoids running actual mutation testing tools and should not be used to write new tests.

How it works

1. Detect language and load extension by calling the language-specific extension guidance file via the extension mechanism.
2. Gather production code and test code pairs, establishing which tests exercise which production methods through call chains and setup helpers.
3. Identify mutation points by examining production code and applying a mutation catalog that covers boundary, boolean/logic, return value, exception, arithmetic, and null-check mutations.
4. Evaluate each mutation against tests by determining which tests cover the mutated line, whether assertions would detect the change, and classifying the mutation as Killed, Survived, No coverage, or Equivalent.
5. If possible, verify survivors by running the test suite, applying each mutation, and re-running the relevant tests to confirm the outcome; otherwise report unverified (static reasoning).
6. Calibrate findings with rules about not flagging trivial code, considering defensive depth, and avoiding false positives; and generate a structured report.

When to use it

• When asked whether user tests would catch a bug if code changes occur.
• To find weak or shallow tests and evaluate test effectiveness beyond coverage.
• When mutation analysis or mutation testing analysis is requested.
• As a pre-check step on freshly generated tests in a test-generation workflow.

What it can touch

The skill references and coordinates with language extensions and production/test code, and uses the mutation catalog to identify and annotate mutation points. It does not itself execute mutation testing tools within this description, and relies on test execution tools available in the environment when verification is possible.

Caveats

• Do not use for writing new tests or for detecting anti-patterns or measuring assertion diversity.
• Do not run actual mutation testing frameworks as part of this skill’s core workflow; if such tools are run, they should be treated as unverified steps and clearly labeled as such.
• If the suite cannot be built or run, findings are labeled unverified (static reasoning) and downgraded in confidence.