What problem does it solve? End-to-end test suites often become flaky, slow, and untrusted, forcing teams to retry builds until green and ignore real failures. This Skill engineers deterministic browser tests that own their data, wait on conditions instead of timers, and produce debuggable failure artifacts. ## Core Features & Use Cases - Deterministic Test Authoring: Writes Playwright and Cypress tests using role-based selectors, API-driven setup, and condition-based waits with zero hard sleeps. - CI Pipeline Engineering: Configures sharded parallel execution, trace-on-retry forensics, and merge-blocking quality gates in GitHub Actions. - Flake Triage & Root-Cause Analysis: Diagnoses flaky tests by signature (timing races, shared state, animation timing) and fixes root causes rather than deleting tests. - Use Case: A team inherits a 40-minute suite at 70% pass rate. Use this Skill to rebuild it with worker-scoped auth fixtures, per-test data factories, and sharded CI, producing an 8-minute merge-blocking suite. ## Quick Start Ask the agent to write a deterministic Playwright test for your checkout flow with API-based setup and role selectors, then wire it into a sharded CI pipeline.