rdg-abgrenzung

Pass

Audited by Gen Agent Trust Hub on Sep 6, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is designed for legal research and drafting regarding the German Rechtsdienstleistungsgesetz (RDG). It references official government websites (gesetze-im-internet.de) and established legal documentation services (dejure.org) for statutes and case law, which are well-known and trusted sources.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests user-provided fact patterns for analysis. While this presents a surface for indirect prompt injection, the skill's lack of dangerous capabilities (such as network exfiltration or file writing) ensures that this surface is not exploitable. The risk is assessed as minimal.
  • [COMMAND_EXECUTION]: The test file documentation includes a command for running a local evaluation script (python ../../../scripts/eval.py). This is a standard internal testing procedure within the project repository and does not involve remote code execution or untrusted sources.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 6, 2026, 11:21 AM
Security Audit — agent-trust-hub — rdg-abgrenzung