dsv-bussgeldverteidigung-art-83
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from legal case files, administrative acts, and evidence (SKILL.md), which lacks boundary markers or sanitization. This creates a surface for potential hidden instructions, though the skill's text-focused capabilities limit the overall risk.
- [SAFE]: The skill references established legal research platforms including gesetze-im-internet.de, dejure.org, and openJur, which are recognized and reputable services for legal information.
Audit Metadata