data-act-nutzerdatenzugang

Pass

Audited by Gen Agent Trust Hub on Sep 6, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied text ("Wortlaut des Zugangsverlangens") that is analyzed by agents to perform legal drafting. The absence of strict boundary markers or sanitization logic creates a surface where an adversary could provide a legal request containing instructions designed to bias the AI's legal analysis or override its internal logic.
  • Ingestion points: User-provided legal request text defined as input in SKILL.md.
  • Boundary markers: The instructions do not specify the use of delimiters or 'ignore' instructions for the ingested text.
  • Capability inventory: The skill utilizes a local Python script for calculations and accesses legal references.
  • Sanitization: No sanitization or validation of the user-supplied text is described.
  • [COMMAND_EXECUTION]: The skill documentation includes instructions for running a local Python module (scripts.legal_calc.cli) to perform automated date calculations based on input parameters.
  • [EXTERNAL_DOWNLOADS]: The skill provides links to official and well-known legal repositories, including EUR-Lex and the German federal justice portal (gesetze-im-internet.de), to support its legal research functions.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 6, 2026, 11:21 AM
Security Audit — agent-trust-hub — data-act-nutzerdatenzugang