rechteclearing-ki-dataset-musik
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill consists of instructional text for legal workflows and does not include any executable code, scripts, or persistent configurations.
- [EXTERNAL_DOWNLOADS]: The skill mentions official legal repositories (e.g., gesetze-im-internet.de, eur-lex.europa.eu) for verifying law and case information. These are established, well-known public services.
- [INDIRECT_PROMPT_INJECTION]: Ingestion points: The skill is designed to process user-provided contracts, metadata, and document files. Boundary markers: No specific delimiters are specified for untrusted data. Capability inventory: No high-privilege tools, subprocesses, or network-write commands are present. Sanitization: No sanitization logic is provided. Assessment: The risk is safe due to the lack of exploitable capabilities within the skill definition.
Audit Metadata