What problem does it solve? Teams often write code before agreeing on observable behavior, leading to misaligned features, untestable specs, and implementation details leaking into requirements. This Skill enforces a disciplined Acceptance Test Driven Development workflow where Given/When/Then specs come before code, and both acceptance and unit test streams must pass. ## Core Features & Use Cases - Gherkin Spec Authoring: Guides writing standard Gherkin spec.md files in pure domain language, with strict rules preventing implementation leakage (no class names, endpoints, or database references). - Test Pipeline Generation: Parses specs into a JSON IR via dae_gherkin.py and invokes a pipeline-builder agent to generate executable acceptance tests for frameworks like pytest, Jest, JUnit, Go testing, or RSpec. - Red-Green TDD Loop: Runs acceptance tests to confirm they fail, then drives implementation with unit tests until both streams pass, with optional impact analysis for faster iteration. - Use Case: When asked to build a new feature, the Skill walks through clarifying questions, co-authors approved Gherkin scenarios, generates a runnable test pipeline, and implements the feature via TDD until all scenarios pass. ## Quick Start Ask the AI to build a feature using ATDD, for example: use ATDD to implement user registration with acceptance specs first.