numerical-check

Pass

Audited by Gen Agent Trust Hub on Aug 24, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to use the Bash tool to run Python scripts via uv run, which is necessary for its primary purpose of mathematical falsification.
  • [PROMPT_INJECTION]: The skill ingests mathematical claims to generate evaluation code.
  • Ingestion points: Mathematical conjectures provided to the agent in SKILL.md.
  • Boundary markers: None explicitly defined.
  • Capability inventory: Execution of Python code via the Bash tool.
  • Sanitization: The agent is provided with a rigid script skeleton to constrain the generated logic.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 24, 2026, 06:17 AM
Security Audit — agent-trust-hub — numerical-check