linyuebanzi-image-gen
Pass
Audited by Gen Agent Trust Hub on Jun 16, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSDATA_EXFILTRATION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill downloads generated image files from remote API providers (MuleRun, APImart, and Atlas Cloud) to the local filesystem using the
urllib.requestlibrary. - [DATA_EXFILTRATION]: Sends user-supplied text prompts and image URLs to external image generation API endpoints. This is the primary function of the skill and is performed via standard HTTP POST requests with API keys sourced from environment variables.
- [PROMPT_INJECTION]: As an execution layer, the skill is susceptible to indirect prompt injection if the calling skill provides malicious prompts. It includes a mitigation feature in
scripts/generate.pythat allows checking prompts against a user-defined blocklist of forbidden terms.
Audit Metadata