What problem does it solve? Manually clicking through a web app to find bugs is slow and easy to skip, and bugs that are found still need to be fixed, committed, and re-verified. This Skill automates the full loop: it drives a real browser through your app, catalogs issues by severity, fixes them in source code with atomic commits, and produces before/after evidence with a ship-readiness summary. ## Core Features & Use Cases - Browser-based QA testing: Navigates pages, clicks every interactive element, fills forms, checks console errors, and captures annotated screenshots across three tiers (Quick, Standard, Exhaustive). - Test-and-fix loop: Fixes bugs directly in source code with one atomic commit per fix, then re-verifies each fix with before/after screenshot evidence. - Diff-aware mode: On a feature branch with no URL given, analyzes the git diff to identify affected pages and routes, then tests exactly what changed. - Test framework bootstrap: Detects your runtime (Node, Ruby, Python, Go, Rust, PHP, Elixir), installs a test framework, writes real regression tests, and sets up CI. - Use Case: You just finished a billing feature on a branch. Run /qa and it detects your local dev server, tests the billing flow end-to-end, finds a broken form validation, fixes it, commits the fix, writes a regression test, and reports a health score improvement from 72 to 94. ## Quick Start Ask the AI to QA test the current feature branch and fix any bugs it finds, or provide a URL like http://localhost:3000 to test a running app.