MCP server that gives Claude direct control of a real browser via Playwright — AI-driven web testing, autonomous test execution, and live failure analysis through natural language
Skill Claude CodeCodex
Execute written browser test scenarios (smoke, regression, or ad-hoc numbered steps) against a live web app with a Playwright browser MCP server, then report a per-step PASS/FAIL table with evidence and a root-cause verdict. Use when asked to "run the smoke suite", "run regression", "run scenario AUTH-1", "test this…
Skill Claude CodeCodex
Turn a browser flow that was just explored or verified (through a Playwright MCP server, a scenario file, or a description) into a committable Playwright Test spec that follows the target repo's own conventions — page objects, fixtures, tags, locator style. Use when asked to "write a test for this", "generate a spec"…