What problem does it solve? Writing end-to-end tests often means guessing selectors, running tests blindly, and manually debugging failures. This Skill automates the full cycle of writing pytest-playwright tests, validating selectors against a real browser, and diagnosing failures until tests pass. ## Core Features & Use Cases - Test Generation: Writes pytest-playwright test files following page object model patterns, conftest fixtures, and locator priority rules from project conventions. - Real Browser Validation: Uses Playwright MCP to verify selectors actually exist on live pages before and after writing tests. - Self-Healing Debug Loop: Diagnoses failures via a three-way check (error output, live browser, frontend source), distinguishes test bugs from app bugs, and auto-repairs changed selectors with an auditable LOCATOR_CHANGES.md log. - Use Case: Ask it to test the event booking flow, and it writes tests/test_booking.py, verifies selectors against the live site, runs pytest, fixes any failures, and reports which business rules are covered. ## Quick Start Ask the agent to generate and validate pytest-playwright E2E tests for the booking cancellation flow.
