media-depth

Pass

Audited by Gen Agent Trust Hub on Jun 27, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill downloads pre-trained model weights from official repositories on the HuggingFace Hub, specifically from the 'depth-anything' and 'Intel' organizations. These are recognized, well-documented models used for computer vision tasks.
  • [COMMAND_EXECUTION]: The scripts/depth.py script uses subprocess.run to call the system's ffmpeg utility. This is used for processing video frames and encoding output videos. The execution is handled safely by passing arguments as a list and using controlled file paths and numerical parameters.
  • [SAFE]: All dependencies are standard Python packages for scientific computing and image processing, and no obfuscation or suspicious network activities were identified.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 27, 2026, 09:53 AM
Security Audit — agent-trust-hub — media-depth