skills/bria-ai/bria-skill/image-utils/Gen Agent Trust Hub

image-utils

Pass

Audited by Gen Agent Trust Hub on Apr 18, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill performs legitimate image processing tasks using the industry-standard Pillow library. All functions (resizing, cropping, adjustments) are deterministic and operate on pixel data as described.\n- [SAFE]: Network activity is restricted to fetching image data via the load_from_url method, which utilizes the requests library with appropriate timeout settings.\n- [SAFE]: File system operations are scoped to loading and saving images, with directory creation limited to the target output path.\n- [SAFE]: No evidence of prompt injection, obfuscation, or unauthorized data exfiltration was found. The skill's behavior is consistent with its stated purpose of assisting in image post-processing.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 18, 2026, 11:55 AM