senior-data-scientist

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • [Prompt Injection] (SAFE): No patterns for overriding instructions or bypassing safety guidelines were found.
  • [Data Exposure & Exfiltration] (SAFE): No sensitive file access or network-based exfiltration logic exists in the provided scripts.
  • [Obfuscation] (SAFE): All code and documentation are in plain text with no hidden or encoded segments.
  • [Unverifiable Dependencies & Remote Code Execution] (SAFE): No remote script downloads or dynamic execution of external code were detected.
  • [Privilege Escalation] (SAFE): The skill does not attempt to gain administrative or root access.
  • [Persistence Mechanisms] (SAFE): No mechanisms for maintaining cross-session access were found.
  • [Metadata Poisoning] (SAFE): Skill metadata is consistent with its stated purpose and lacks deceptive instructions.
  • [Indirect Prompt Injection] (SAFE): The skill has no interface that would allow untrusted data to influence agent behavior via prompt interpolation.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 04:36 PM