skills/klotzkette/claude-fuer-deutsches-recht/freelancer-misclassification-ip-risk/Gen Agent Trust Hub
freelancer-misclassification-ip-risk
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data such as case facts, contracts, and legal documents to perform its analysis.
- Ingestion points: Case details, intake question responses, and documents provided by the user (identified in 'Workflow' and 'Mindest-Output' sections).
- Boundary markers: None explicitly defined in the skill instructions to separate user data from instructions.
- Capability inventory: The skill is limited to text generation (memos, matrices, briefings). No network-writing, file-system-writing, or code-execution capabilities were identified.
- Sanitization: No specific sanitization or validation of the input content is mentioned.
Audit Metadata