Qualmly

About Qualmly

Qualmly is an AI reviewer aimed specifically at applications shipped out of Lovable, Bolt, v0, Cursor, GitHub Copilot, Claude Code, Windsurf, Replit and Webflow — the class of app where nobody on the team is necessarily an engineer and the failure modes are consistent enough to check for by name. You paste an app URL, or the raw HTML when the app sits behind a login or blocks crawlers, pick the app type, and it returns findings across eight test categories in plain English in about thirty seconds. The Vibe-Coded App preset targets the specific patterns AI code generators produce: missing Supabase row-level security, secret keys leaked into the client bundle, unauthenticated endpoints, exposed admin routes and boilerplate CORS/CSRF mistakes. There is also an opt-in active Supabase RLS probe that goes past passive recon and actually attempts read-only queries against the Supabase project found in the bundle to confirm whether RLS is enforced — capped at ten requests, 100 ms apart, read-only, and gated behind an explicit ownership attestation with a clear warning that running it against someone else's app may breach their terms of service or unauthorised-access law. Beyond the scanner there is a code-review surface where you paste JavaScript, React, HTML or CSS and get a UAT and best-practice audit with one-click inserts, plus a monitoring tab. It runs as a single HTML file against your own Claude key, and the QA check is free.

Key Features

Tags

More AI Security & Testing tools