puda-machine-vision-validation

Pass

Audited by Gen Agent Trust Hub on Jul 21, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is designed as a safety gate to prevent unauthorized or unsafe machine motion. It emphasizes environment isolation, provenance tracking (SHA-256 hashes of images), and explicit user approval for any state-changing actions.
  • [COMMAND_EXECUTION]: The instructions involve running platform-specific commands such as puda env current to determine the active context. This is a standard part of the environment-aware workflow and poses no inherent security risk.
  • [EXTERNAL_DOWNLOADS]: The skill suggests installing environment-specific dependencies using puda skills install pudap/<environment>-skills. These resources are from the same vendor ('pudap') and are part of the documented multi-site architecture.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 21, 2026, 05:32 AM
Security Audit — agent-trust-hub — puda-machine-vision-validation