simple-math

Pass

Audited by Gen Agent Trust Hub on Feb 19, 2026

Risk Level: SAFE
Full Analysis
  • Prompt Injection (SAFE): No instructions found that attempt to override agent behavior or bypass safety filters.
  • Data Exposure & Exfiltration (SAFE): No sensitive file access or network operations detected. The code only performs local mathematical calculations.
  • Obfuscation (SAFE): No encoded strings, zero-width characters, or hidden content identified in the scripts or metadata.
  • Unverifiable Dependencies & Remote Code Execution (SAFE): No external dependencies or remote script downloads are present. The implementation relies entirely on the Python standard library.
  • Indirect Prompt Injection (SAFE): While the skill processes user-provided inputs ('a', 'b', and 'op'), it strictly validates them by casting to floats and checking against a whitelist of operators. There is no execution sink (like eval/exec) for malicious strings to be processed.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 19, 2026, 05:24 PM