test-quality

Map expected behavior to existing tests, then identify uncovered critical paths, negative scenarios, and boundary conditions. This Skill performs that mapping and lists concrete missing scenarios tied to specific risks.

Flaky tests usually come from non-deterministic setup, shared mutable state, time or network dependencies, and poor isolation between cases. The review checks setup and teardown determinism and flags shared-state patterns as stability risks.

Weak assertions only check status codes, response lengths, or that no exception occurred, without verifying actual behavior. Strong assertions validate meaningful outputs, include boundary and negative cases, and produce failure messages with triage value.

Skip it for pure feature implementation, refactors without test-quality intent, or general code reviews that are not test-focused. It is designed specifically for assessing and improving test suites, not writing product code.

Recommendations are ranked by risk reduction versus implementation cost and split into must-now and next tiers. Each recommendation is tied to a concrete risk, and the output defines acceptance criteria for the upgraded suite.