aliyun-qwen-multimodal-embedding

Pass

Audited by Gen Agent Trust Hub on Apr 29, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides a legitimate interface for generating JSON payloads required for Alibaba Cloud Model Studio multimodal embedding models.
  • [DATA_EXPOSURE]: The skill documentation correctly guides users to manage their authentication credentials (DASHSCOPE_API_KEY) via environment variables or local configuration files, which is consistent with secure development practices.
  • [COMMAND_EXECUTION]: The commands included in the documentation for validation and usage (e.g., mkdir, py_compile) are standard, benign development operations and pose no risk.
  • [INDIRECT_PROMPT_INJECTION]: The script prepare_multimodal_embedding_request.py processes user-supplied strings and URLs. However, it uses the standard json.dumps library to safely serialize this data into a JSON structure, preventing command or script injection within the generated payload.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 29, 2026, 02:27 AM