wechat-tech-writer
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION] (LOW): The skill contains an indirect prompt injection surface as it passes user-supplied prompts directly to third-party image generation APIs.\n
- Ingestion points:
scripts/generate_image.pyvia the--promptargument.\n - Boundary markers: Absent. The prompt is passed to the models as a raw string.\n
- Capability inventory: Network operations to Google and OpenAI APIs.\n
- Sanitization: None. No filtering or encoding of user-provided prompt strings.\n- [DATA_EXPOSURE] (LOW): The scripts
generate_cover_optimized.pyandgenerate_temp.pyinclude hardcoded Windows absolute file paths (e.g.,G:\git_pull\...). This exposes the developer's local file structure and limits script portability.\n- [EXTERNAL_DOWNLOADS] (SAFE): The skill utilizes legitimate, well-known libraries (google-genai,requests) and interacts exclusively with official API endpoints for Google and OpenAI.\n- [CREDENTIALS_UNSAFE] (SAFE): API keys are managed through environment variables rather than being hardcoded in scripts or documentation, aligning with security best practices.
Audit Metadata