gpt-image-2-product-retouch
Pass
Audited by Gen Agent Trust Hub on Aug 12, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is a legitimate API wrapper for an image processing service. It follows security best practices by ensuring its configuration file, which stores API keys, is restricted to owner-only access (mode 0600).
- [EXTERNAL_DOWNLOADS]: The script downloads processed results from the service's API at ai-hive.iclip.cn. This behavior is consistent with the skill's purpose of retrieving retouched product images.
- [DATA_EXFILTRATION]: The skill transmits user-selected image files to the AI Hive endpoint for processing. This data transmission is the core functionality of the retouching service and is performed over secure HTTPS connections.
- [COMMAND_EXECUTION]: The Python script provides a CLI for task management and media uploads. It does not perform any arbitrary or unauthorized command execution on the host system.
Audit Metadata