skills/klotzkette/claude-fuer-deutsches-recht/bearbeitung-umgestaltung-beweisforensik/Gen Agent Trust Hub
bearbeitung-umgestaltung-beweisforensik
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external, untrusted legal data which creates a potential surface for indirect prompt injection attacks.
- Ingestion points: The skill instructs the agent to gather and analyze external documents including administrative acts (Verwaltungsakte), contracts (Vertragsurkunden), legal briefs (Schriftsätze), and digital metadata (SKILL.md).
- Boundary markers: The instructions lack explicit delimiters or specific 'ignore embedded instructions' warnings for the processing of these external documents.
- Capability inventory: The skill is focused on analytical outputs (matrices, memos, and risk assessments); it does not exhibit dangerous capabilities such as arbitrary command execution, persistent file-writing, or network exfiltration to untrusted domains.
- Sanitization: There is no evidence of automated sanitization or filtering logic for the content of processed documents.
Audit Metadata