Free Security Scan for AI Apps
Catch the security issues that vibe-coded apps actually ship with. Passive checks only — no probing, no signups, no catch.
Need a deeper, code-level scan?
The free scan above checks what your site exposes at the network layer. Scanner Pro reads your actual source code — finds hardcoded secrets across every file, dependency manifest issues, and AI-specific risk patterns the surface scan can't see.
- Hardcoded secret patterns across the entire repo (~30 secret types)
- Dependency manifest analysis (npm, pip, Cargo, Go, Maven)
- AI/LLM code patterns mapped to OWASP LLM Top 10 categories
- Public OR private GitHub repos (Personal Access Token supported, never stored)
- 3 deep repo scans per month, cancel any time
What we check
Security headers
CSP, HSTS, X-Frame-Options, X-Content-Type-Options, Referrer-Policy, Permissions-Policy.
TLS / HTTPS
Certificate validity, protocol versions, common misconfigurations.
Cookie security
Secure, HttpOnly, and SameSite flags on cookies served to visitors.
Mixed content
HTTP resources pulled into HTTPS pages — common cause of browser warnings.
Client-side secret exposure
API keys and tokens accidentally shipped in returned HTML or JavaScript.
AI feature risk profile
Detection of chat/AI features and the categories of injection risk that apply. Informational, not a vulnerability assertion.
Passive only. We fetch your URL the way any browser would. We do not probe paths you don't own, attempt unauthorized access, or send adversarial inputs. For deeper checks (hardcoded secrets, dependency vulnerabilities, code-level AI risk patterns), Scanner Pro reads your code with explicit access — your Personal Access Token, never stored.