seo-image-gen

Pass

Audited by Gen Agent Trust Hub on Jun 16, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The scripts/setup_mcp.py script configures the agent platform to download and execute the @ycse/nanobanana-mcp package from the NPM registry using npx. This package provides the core Model Context Protocol (MCP) server used for image generation.
  • [COMMAND_EXECUTION]: The skill uses several Python scripts (generate.py, edit.py, cost_tracker.py, presets.py, batch.py) to manage API calls, brand presets, and usage logs. The presets.py script includes path sanitization logic to prevent directory traversal during preset management.
  • [COMMAND_EXECUTION]: Documentation within references/post-processing.md provides instructions for using local system-level tools such as ImageMagick (magick) and FFmpeg for image manipulation tasks like resizing, background removal, and format conversion.
  • [SAFE]: The skill communicates with official Google API endpoints (generativelanguage.googleapis.com) to perform its image generation and editing tasks. This involves transmitting user-provided prompts and image data to a well-known service as part of the primary functionality.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 16, 2026, 02:37 AM
Security Audit — agent-trust-hub — seo-image-gen