skills/klotzkette/claude-fuer-deutsches-recht/akteneinsicht-strafrecht-auswerten/Gen Agent Trust Hub
akteneinsicht-strafrecht-auswerten
Pass
Audited by Gen Agent Trust Hub on Sep 3, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingest and processes untrusted external content in the form of PDF criminal records. This creates a surface where instructions embedded in the documents could potentially influence the agent's behavior. \n
- Ingestion points: User-provided OCR-processed PDFs of legal case files (SKILL.md). \n
- Boundary markers: None. The skill does not define specific delimiters or instructions to ignore embedded commands within the processed text. \n
- Capability inventory: Limited to analytical processing and markdown file generation; no high-risk tools like shell access are invoked. \n
- Sanitization: No measures for filtering or escaping the text from the source files are specified. \n- [EXTERNAL_DOWNLOADS]: The skill requires the agent to verify legal citations using live data from external repositories. \n
- Sources: Official and established legal information portals including gesetze-im-internet.de, dejure.org, and openJur. \n
- Context: Verification of statutes and case law to ensure analytical accuracy against authoritative sources.
Audit Metadata