What problem does it solve? After code validation, teams need a disciplined way to prove that a feature behaves correctly against its tests, acceptance criteria, and planned end-to-end slice, with failures classified instead of silently ignored. ## Core Features & Use Cases - Structured Test Execution: Runs focused feature tests first, then nearby regression tests, then integration or component tests required by the design, and broader suites only when feasible. - Failure Classification: Maps every failure to a category such as IMPLEMENTATION_DEFECT, TEST_DEFECT, ENVIRONMENT_DEFECT, or DESIGN_DEFECT before any retry or routing backward. - Evidence Generation: Writes a tests.md evidence file from a template, bound to the exact integration HEAD SHA, and moves the feature state to tested. - Use Case: After sdlc-tdd creates tests for a feature, invoke this skill to run them from the verified integration worktree, prove the planned end-to-end slice, and produce auditable pass/fail evidence. ## Quick Start Ask the Agentic SDLC coordinator to run the unit tests phase for the current feature so the planned tests execute and evidence is recorded.
