media-upscale

Pass

Audited by Gen Agent Trust Hub on Jun 27, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The script scripts/upscale.py facilitates media processing by executing external binaries such as ffmpeg, realesrgan-ncnn-vulkan, and realcugan-ncnn-vulkan. These operations are implemented using subprocess.run() with list-based arguments, which effectively mitigates command injection risks by avoiding shell interpretation.
  • [EXTERNAL_DOWNLOADS]: The skill documentation and the install subcommand provide users with instructions for downloading super-resolution models and libraries from well-known public repositories on GitHub (e.g., xinntao/Real-ESRGAN, TencentARC/GFPGAN). The skill does not perform any unattended background downloads or installations; it instead empowers the user to manage dependencies through standard package managers or manual verification.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 27, 2026, 09:54 AM
Security Audit — agent-trust-hub — media-upscale