Val is the QA agent you talk to inside Cursor, Claude Code, or Codex. Ask your AI to check your build, Val opens a real…
Val is the QA agent you talk to inside Cursor, Claude Code, or Codex. Ask your AI to check your build, Val opens a real Chromium browser via Playwright, walks every flow, runs the full Lighthouse and axe-core audits, captures the network, and returns a structured report. Cursor reads it, edits your code, Val re-runs to verify. You ship to green without bothering anyone. Tests deployed apps and localhost. Walks any signup, checkout, or paid funnel: Val auto-fills the Stripe 4242 test card in test mode, and refuses on live mode to avoid real charges. For Paddle, LemonSqueezy, or Gumroad, Val flags the processor and recommends switching to Stripe test mode. Logs into authenticated apps with credentials you pass it and persists the session, so every subsequent Val call runs as the authenticated user. What Val catches: broken links and 4xx/5xx responses, uncaught JS errors and console errors, broken images, horizontal layout overflow on mobile, Cumulative Layout Shift, accessibility violations (full WCAG 2.0/2.1 A + AA via axe-core), bugs hidden in HTTP 200 responses with error:true in the body, dead buttons, broken forms, mobile-only and desktop-only bugs. Posts findings as PR comments on GitHub: pass a PR URL, Val reads the diff, infers the touched routes, tests them on the preview deployment, posts one PR comment updated in place across re-runs. Multi-device testing (desktop, iPhone 14, Pixel 7).
arthur-visionartise
mcp
free
Others in the same category, ranked by how often they are opened.
