sekretariats-tagesbrief
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data which could contain malicious instructions.
- Ingestion points: Processes incoming mail (e.g., 'Posteingang vom Vortag'), external evidence files, and court decisions retrieved from official sources.
- Boundary markers: Absent. The instructions do not specify delimiters or warnings to ignore instructions embedded within the legal documents or incoming mail.
- Capability inventory: The skill generates legal drafts, memos, and briefings. It also writes files to the local configuration directory (
~/.claude/plugins/config/claude-fuer-deutsches-recht/). - Sanitization: Absent. There is no mention of sanitizing or escaping content from external documents before including them in the generated briefings or drafts.
- [EXTERNAL_DOWNLOADS]: The skill instructs the agent to perform live checks of legal norms and court decisions from external sources.
- Evidence: The workflow requires 'Live-Check für tagesaktuelle Normänderungen' and verifying decisions via 'amtliche oder frei zugängliche Quelle'. While these are directed at official services, they represent runtime retrieval of external content.
Audit Metadata