What problem does it solve? Automated unit and integration tests miss visual bugs, broken interaction flows, accessibility violations, and edge-case failures that only appear when a real user drives a live application. This Skill performs hands-on QA against a running app using browser automation, producing a structured severity-classified report. ## Core Features & Use Cases - Browser-driven testing: Controls a real browser via Playwright to run smoke tests, exploratory break-things testing, visual review, accessibility checks, form boundary testing, and network inspection. - Full application lifecycle handling: Detects the framework (Vite, Next.js, Django, Flask, FastAPI, Go), starts the dev server, waits for HTTP readiness, and cleans up processes afterward. - Structured QA report: Classifies every finding as P0-P3 with reproduction steps, screenshots, and documented passed checks. - Use Case: After building a new settings page, ask the agent to QA the running app — it will start the dev server, probe form fields with boundary and injection inputs, capture console errors and screenshots across breakpoints, and return a prioritized findings report. ## Quick Start Ask the agent to QA test the running application at localhost:3000 and report any bugs, accessibility issues, or visual problems it finds.
