image-generation

Fail

Audited by Gen Agent Trust Hub on Feb 25, 2026

Risk Level: CRITICAL
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill utilizes the 'z-ai-web-dev-sdk' package and the 'z-ai' CLI tool. Both resources are legitimate components provided by the vendor 'jjyaoao' and are used for their intended purpose of image generation.\n- [PROMPT_INJECTION]: The skill processes user-provided text prompts to generate images, which presents a surface for indirect prompt injection. While no malicious overrides were found, the skill lacks input sanitization for these prompts.\n
  • Ingestion points: The prompt parameter in generateImage, generateImageWithSize, and batch processing functions within SKILL.md and scripts/image-generation.ts.\n
  • Boundary markers: None. The prompts are interpolated directly into the API request payload.\n
  • Capability inventory: The skill performs local file system writes via fs.writeFileSync to save generated images.\n
  • Sanitization: No validation or filtering of prompt content is implemented before processing.
Recommendations
  • Contains 1 malicious URL(s) - DO NOT USE
Audit Metadata
Risk Level
CRITICAL
Analyzed
Feb 25, 2026, 03:56 AM