ian-xiaohei-cat-illustrations
Pass
Audited by Gen Agent Trust Hub on Jul 27, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill uses instructional constraints to maintain artistic style and character integrity (e.g., '禁止 PPT', '绝对不要'). These are functional requirements for the task and do not attempt to bypass underlying AI safety filters.
- [DATA_EXFILTRATION]: No unauthorized data access or external transmission patterns were detected. The skill uses local directory paths for organizing generated assets, which is standard behavior for creative workflows.
- [REMOTE_CODE_EXECUTION]: The skill does not include any external dependencies, package installations, or script execution patterns. It relies solely on the built-in image generation tool.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data (user-provided articles, links, and documents) to create image generation prompts. While this is an ingestion surface, the risk is mitigated by a very specific and prescriptive prompt template and a mandatory QA checklist that requires high similarity to internal reference images.
- [COMMAND_EXECUTION]: No dangerous system commands or shell executions were found. The workflows are limited to text analysis and image generation tool calls.
Audit Metadata