gemini-imagegen

Pass

Audited by Gen Agent Trust Hub on Mar 30, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill integrates with the official google-genai library and Pillow for image processing. Both are standard, well-known dependencies.
  • [SAFE]: Sensitive credentials are managed securely via the GEMINI_API_KEY environment variable, preventing accidental exposure in logs or code.
  • [SAFE]: File system operations are restricted to common image formats through the PIL (Pillow) library, and file paths are explicitly provided as command-line arguments or interactive inputs.
  • [SAFE]: The skill's instructions and scripts do not contain any obfuscated code, unauthorized network requests, or attempts to bypass agent safety guidelines.
  • [SAFE]: No persistence mechanisms, privilege escalation attempts, or dynamic execution patterns were detected.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 30, 2026, 01:20 AM