skills/klotzkette/claude-fuer-deutsches-recht/akzg-vertraulichkeit-redaction-spezial/Gen Agent Trust Hub
akzg-vertraulichkeit-redaction-spezial
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 ingest and process user-provided legal facts and documents to generate structured work products such as memos and drafts.
- Ingestion points: The skill requests case facts, evidence, and legal documents (e.g., 'Sachverhalt', 'Unterlagen', 'Beweismittel') in the 'Arbeitsweg' and 'Fallweichen' sections.
- Boundary markers: There are no explicit boundary markers or instructions to the agent to disregard instructions potentially embedded within user-supplied legal text.
- Capability inventory: The skill generates text-based work products. It does not contain scripts or commands that access the file system or execute arbitrary code.
- Sanitization: No input validation or sanitization mechanisms are defined for the case-related data processed at runtime.
Audit Metadata