image-prompt-generator

Pass

Audited by Gen Agent Trust Hub on Feb 16, 2026

Risk Level: LOW
Full Analysis
  • [EXTERNAL_DOWNLOADS] (LOW): The script scripts/generate_image.py depends on google-genai and pillow. These are widely used and reputable libraries for API interaction and image manipulation.
  • [COMMAND_EXECUTION] (LOW): The tool is a command-line script designed to be executed by an agent or user to generate images. It uses standard argparse for input handling.
  • [DATA_EXFILTRATION] (SAFE): Sensitive information (API keys) is handled securely via environment variables (GEMINI_API_KEY or GOOGLE_API_KEY). No hardcoded credentials or unauthorized network calls were found.
  • [PROMPT_INJECTION] (SAFE): The style templates in the references/ directory provide structured guidance for image generation. They do not contain instructions to bypass AI safety filters or leak system prompts.
  • [INDIRECT_PROMPT_INJECTION] (LOW): While the skill (specifically in schematic.md) describes processing external data such as Telegram captions, its capability is limited to image generation, which poses a low risk of influencing agent logic or performing unauthorized actions.
Audit Metadata
Risk Level
LOW
Analyzed
Feb 16, 2026, 06:08 AM