812-regulations-eu-product-liability-directive
Pass
Audited by Gen Agent Trust Hub on Aug 3, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is purely instructional and advisory, containing no executable code, malicious commands, or unauthorized access patterns.
- [DATA_EXFILTRATION]: Incorporates robust security practices by mandating the redaction of sensitive information, including API keys, credentials, and personal data, throughout the evidence collection and reporting phases.
- [PROMPT_INJECTION]: Evaluates the ingestion of project-specific data while providing clear guidelines for using trusted evidence and sanitizing inputs, effectively managing the surface for indirect prompt injection.
- [EXTERNAL_DOWNLOADS]: References official regulatory sources from the European Union (eur-lex.europa.eu) and explicitly restricts the agent from fetching external content at runtime, ensuring dependency on verified, bundled reference materials.
Audit Metadata