What problem does it solve? Writing tests after implementation often produces brittle tests coupled to internal structure, while writing all tests upfront produces tests of imagined behavior. This Skill enforces a disciplined red-green-refactor loop so tests verify observable behavior through public interfaces and survive refactors. ## Core Features & Use Cases - Red-Green-Refactor Workflow: Drives one tracer-bullet test at a time through failing test, minimal implementation, and refactoring phases. - Test Quality Guardrails: Defines anti-patterns such as tautological assertions, horizontal slicing, and mocking internal collaborators, with worked good/bad examples. - Five-Layer Test Taxonomy: Provides a shared vocabulary (acceptance, contract, property-based, integration, unit) plus a decision matrix mapping change types to required test layers. - Use Case: When asked to implement a checkout feature with TDD, the agent confirms the interface and behaviors to test, writes one failing test, implements the minimal code to pass, and repeats per behavior before refactoring. ## Quick Start Ask the agent to build the next feature or bug fix using TDD with the red-green-refactor loop, confirming the public interface and behaviors to test first.