image-upscaling

Pass

Audited by Gen Agent Trust Hub on May 5, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill utilizes the vendor's official CLI tool, belt, to perform image upscaling tasks through their API.
  • [SAFE]: All external URLs and repository references point to the author's official domains (inference.sh) and GitHub organization (inference-sh), which is consistent with the vendor's identity.
  • [SAFE]: The skill explicitly restricts tool usage to the belt command via the allowed-tools frontmatter, implementing a principle of least privilege.
  • [SAFE]: No malicious patterns such as prompt injection, credential harvesting, or unauthorized remote code execution were detected in the instructions or examples provided.
Audit Metadata
Risk Level
SAFE
Analyzed
May 5, 2026, 01:17 AM
Security Audit — agent-trust-hub — image-upscaling