abmahnung-und-cease-desist
Pass
Audited by Gen Agent Trust Hub on Sep 3, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external legal documents which could contain adversarial instructions.
- Ingestion points: The skill instructions in SKILL.md direct the agent to analyze "Verwaltungsakte, Vertragsurkunden, Schriftsätze, Bescheide, Protokolle, Sachverständigengutachten".
- Boundary markers: No specific delimiters or safety instructions are provided to help the agent distinguish between its instructions and the content of the documents.
- Capability inventory: The skill does not define any capabilities for network access, file system modification, or command execution.
- Sanitization: No methods for filtering or sanitizing document content are mentioned.
Audit Metadata