agb-entwurf-kaltstart
Pass
Audited by Gen Agent Trust Hub on Sep 3, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to analyze external legal documents provided by the user, which serves as an ingestion point for untrusted data. Instructions hidden within these documents could potentially influence agent behavior.
- Ingestion points: Untrusted material is analyzed in SKILL.md ("Wenn Material vorliegt, lies es zuerst").
- Boundary markers: The skill does not define specific delimiters or instructions for the agent to separate user-provided content from instructions.
- Capability inventory: The agent's capabilities are restricted to text analysis, drafting, and reference checking; no file-writing or network exfiltration tools are authorized.
- Sanitization: No sanitization or input validation steps are included in the workflow.
- [SAFE]: The skill references official legal norms through well-known public resources ("Gesetze im Internet") to ensure accuracy in its legal assessments.
Audit Metadata