Hi HN! I built a GitHub Action that automatically scans AI/LLM endpoints for security vulnerabilities on every push/PR.
Why? Most teams ship AI features without security testing. This action catches prompt injection, jailbreaks, and data leakage before they hit production.
How it works:
- Add 5 lines of YAML to your workflow
- Scans run automatically on push/PR
- PRs get blocked if critical vulns are found
- Full report with remediation steps
Free tier: 5 scans/month. 650+ attack vectors.
Built this because I do AI red teaming professionally (OSCP+, C-AI/MLPen). Happy to answer questions!
xsourcesec•3h ago
Why? Most teams ship AI features without security testing. This action catches prompt injection, jailbreaks, and data leakage before they hit production.
How it works: - Add 5 lines of YAML to your workflow - Scans run automatically on push/PR - PRs get blocked if critical vulns are found - Full report with remediation steps
Free tier: 5 scans/month. 650+ attack vectors.
Built this because I do AI red teaming professionally (OSCP+, C-AI/MLPen). Happy to answer questions!