scientific-reasoning

Pass

Audited by Gen Agent Trust Hub on Apr 6, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is entirely instructional, focusing on mathematical and physical problem-solving protocols without any executable scripts or tools.
  • [SAFE]: No network operations, file system access, or credential usage were detected.
  • [SAFE]: References to external research datasets (MATH Dataset, PhysReason) are used for context and do not involve remote code execution or automated downloads.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 6, 2026, 11:30 PM
Security Audit — agent-trust-hub — scientific-reasoning