agrar-wolfsschaden-spezial
Pass
Audited by Gen Agent Trust Hub on Sep 3, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill serves a legitimate professional purpose, providing guidance for legal assessments related to wolf damage compensation. It references well-known and trusted legal research portals such as gesetze-im-internet.de and dejure.org. No malicious command execution, data exfiltration, or obfuscation patterns were detected.- [INDIRECT_PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it is designed to process external case materials and facts supplied by the user. However, the risk is minimal as the agent's capabilities are limited to text generation for human review.
- Ingestion points: Case facts, documents, and client materials (SKILL.md).
- Boundary markers: Not explicitly defined in the instructions for input interpolation.
- Capability inventory: Textual output generation (memos, pleadings); no shell, file-system write, or unauthorized network capabilities are utilized.
- Sanitization: No specific sanitization or filtering of input content is mandated in the prompt instructions.
Audit Metadata