Observed arrival · 2026-11-01
Emerald Red Team wants AI-built apps to break before launch
A security-testing service for apps built with AI coding tools, with findings presented as prompts, files, or pull requests.
- For
- Makers of AI-built apps using Lovable, Bolt, Cursor, or Replit
- Worth noticing
- Its 15 modules include conventional security checks alongside pre-consent tracking, accessibility, legal pages, and fake reviews.
Field notes
The sample scan names concrete findings, including an exposed Supabase service-role key and an unauthenticated orders table, then shows a grade moving from F toward A as fixes land. Its checklist extends beyond application security into matters such as pre-consent tracking and accessibility. The page cites outside research for its claims about vulnerabilities in AI-built sites; those figures are presented by the service, not independently established by the extracted evidence.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
One card from the complete issue