integrationen-simulation-kanzlei
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process external legal documents and evidence, which creates a surface for indirect prompt injection attacks.
- Ingestion points: The 'Arbeitsweg' and 'Triage' sections describe collecting administrative acts, contracts, and other legal documents in
SKILL.md. - Boundary markers: The skill includes instructions to anonymize client data before entry into simulation mode, though explicit text delimiters for external content are not specified.
- Capability inventory: The skill uses templates (
assets/templates/integrationsstatus-und-simulation.md) and instructs the agent to perform live checks of legal norms via tool calls. - Sanitization: The skill explicitly mandates the anonymization of client data as a prerequisite for using the simulation mode.
Audit Metadata