gpt-image-2-image-to-image
Pass
Audited by Gen Agent Trust Hub on Aug 12, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSDATA_EXFILTRATIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill connects to the domain
ai-hive.iclip.cnto facilitate image generation. This network activity is the primary intended function of the skill and is documented in the instructions. - [DATA_EXFILTRATION]: The script
scripts/imagegen.pyreads user-specified local image files and uploads them to the AI Hive API for processing. This data flow is transparent and initiated by user command-line arguments. - [EXTERNAL_DOWNLOADS]: The utility downloads generated image files from the service's result URLs to the user's local
Downloadsfolder. - [COMMAND_EXECUTION]: The initialization process uses the Python
webbrowsermodule to open an API key management page in the user's default browser. - [PROMPT_INJECTION]: The skill accepts user-provided text prompts and images as input for the external AI model. It lacks specific delimiters or sanitization for these inputs, which is characteristic of a direct API wrapper.
- Ingestion points: User-provided prompt strings and local image file paths supplied to the
generatecommand. - Boundary markers: None present; user input is interpolated directly into the JSON payload for the API.
- Capability inventory: Network requests via the
requestslibrary, file system read/write access, and opening a system web browser. - Sanitization: No input filtering or escaping is performed on the prompt text before transmission.
Audit Metadata