datenschutz-hris-und-saas-tools
Pass
Audited by Gen Agent Trust Hub on Aug 21, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill does not contain any suspicious command execution, privilege escalation, or persistence mechanisms. It is entirely instructional and focused on legal and administrative processes.
- [DATA_EXPOSURE_AND_EXFILTRATION]: There are no indicators of data exfiltration or credential exposure. On the contrary, the skill explicitly instructs the agent to limit the distribution of sensitive HR and payroll data to avoid unnecessary disclosure.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests external data such as contracts and legal documents, which constitutes an indirect prompt injection surface. However, it mitigates this by instructing the agent to verify norms live and clarify missing information with the user rather than blindly following document content.
- [COMMAND_EXECUTION]: No shell commands or script execution patterns were identified. The instructions use references to German laws (BGB, DSGVO, etc.) for purely informational and analytical purposes.
Audit Metadata