Strategy and scope¶

R-STRAT-1 - Critical behavior is mapped to tests¶

Rule. Every repo slice must name its critical behavior and map each item to automated tests or an explicit risk exception.

Why. A green test suite is meaningless when nobody knows which behavior it protects. The map tells reviewers whether important flows are covered or merely hoped for.

vcqa. Flag repos with deployable code but no test strategy, no critical-flow list, and no relationship between product/API/package behavior and test files.

References.

• Testing Library Guiding Principles: https://testing-library.com/docs/guiding-principles

R-STRAT-2 - Test layers match the slice shape¶

Rule. Test layers must match what the slice ships: libraries need consumer/API compatibility tests, CLIs need command/exit-code tests, APIs need boundary tests, and UI apps need component plus smoke tests.

Why. A single test layer cannot prove every runtime. Unit-only coverage does not prove deployment wiring, while E2E-only coverage gives slow, vague feedback for deterministic logic bugs.

vcqa. Compare detected stack/slice signals with test directories, package scripts, and workflow commands; flag missing expected layers for the slice type.

References.

• Vitest Guide: https://vitest.dev/guide/
• Playwright Best Practices: https://playwright.dev/docs/best-practices

R-RISK-1 - Untested risk is explicit¶

Rule. Untested critical paths, manual-only checks, generated-code exclusions, and temporarily deferred test layers must be documented with owner, reason, and review date.

Why. Hidden risk becomes permanent debt. Explicit risk lets a reviewer distinguish a deliberate exception from a forgotten gap.

vcqa. Flag broad coverage/test exclusions or missing test layers without nearby documentation naming reason, owner, or review trigger.

References.

• GitHub Actions deployments and environments: https://docs.github.com/en/actions/reference/workflows-and-actions/deployments-and-environments