cv-opencv

Pass

Audited by Gen Agent Trust Hub on May 9, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill facilitates the installation of official OpenCV packages and provides instructions for downloading models from the official OpenCV GitHub repository.
  • [PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection because it processes untrusted media and data files.
  • Ingestion points: The 'cv.py' script processes images, video streams from URLs, and label files.
  • Boundary markers: None present; tool output is returned to the agent context without specific delimiters.
  • Capability inventory: The skill can perform file system writes and access network-based video streams.
  • Sanitization: No validation or escaping is applied to extracted metadata or labels.
  • [COMMAND_EXECUTION]: The 'calibrate' command in 'scripts/cv.py' uses the 'glob.glob' function on user input, allowing for filesystem interaction via wildcards.
Audit Metadata
Risk Level
SAFE
Analyzed
May 9, 2026, 07:09 AM
Security Audit — agent-trust-hub — cv-opencv