AI System Risk Scan
One GitHub URL. One scan. Two outputs: your EU AI Act risk classification and a technical health report covering security posture, supply-chain exposure, and compliance readiness — everything regulators and engineering leads need in a single pass.
EU AI Act Risk Classification
We read your source code — not a checklist. Risk tier (minimal, limited, high, unacceptable) is determined from actual model usage, decision logic, and deployment context found in the repo.
- Risk tier with article citations
- Prohibited-use pattern detection
- Transparency and explainability gaps
- Human oversight requirement flag
Technical Health Report
Security and supply-chain findings your legal team needs to sign off on and your engineering team needs to fix — surfaced from the same repo scan.
- Dependency vulnerability surface
- Supply-chain lockfile integrity check
- Hardcoded secrets and credential patterns
- CI/CD security posture assessment
How the scan works
Fetch
We clone the public repo and index every file — source code, configs, lockfiles, CI definitions
Classify
LLM-powered analysis maps code patterns to EU AI Act articles and known vulnerability signatures
Report
Risk tier, evidence trail, and technical health findings delivered together — one URL, one binder
Scan your AI repo now
Free preview shows risk tier and top findings. Full report and audit pack generation require sign-in.