nano-banana

Pass

Audited by Gen Agent Trust Hub on May 16, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill communicates exclusively with Google's Generative Language API (googleapis.com), which is a well-known and trusted service. This minimizes the risk of data exfiltration to unauthorized parties.- [SAFE]: Sensitive data management is handled securely. The script requires the NANO_BANANA_API_KEY to be provided via an environment variable, preventing the exposure of credentials in source code or configuration files.- [SAFE]: File operations are limited to necessary functional tasks: reading local configuration files, reading user-provided input images for processing, and writing generated images back to the filesystem with version-controlled filenames to prevent accidental data loss.- [SAFE]: No malicious patterns such as obfuscation, dynamic code execution, or unauthorized privilege escalation were identified during the analysis of the Python scripts and accompanying metadata.
Audit Metadata
Risk Level
SAFE
Analyzed
May 16, 2026, 12:01 AM
Security Audit — agent-trust-hub — nano-banana