What problem does it solve? Writing end-to-end browser tests that stay aligned with product definitions is tedious and error-prone. This Skill reads user stories and acceptance criteria directly from DefProd and generates one test file per UI story, keeping tests traceable to the product definition. ## Core Features & Use Cases - Story-driven test generation: Creates one Playwright or Cypress spec file per in-scope user story, with test cases mapped to individual acceptance criteria. - Surface filtering: Selects only stories whose surface is ui (explicit or inferred from the story-key prefix), skipping API, MCP, and CLI stories for sibling skills. - Testability analysis and coverage reporting: Classifies acceptance criteria as testable or untestable, runs the generated tests, fixes test bugs, and reports a per-story coverage summary. - Use Case: Given a DefProd product with a BILLING area containing UI stories, run the skill to produce Playwright specs in your e2e directory, execute them, and receive a coverage table flagging untestable criteria. ## Quick Start Ask your AI agent to run /defprod-create-ui-tests with your product name and area key to generate and run browser tests for that area's UI stories.