ai-ecommerce-expert-seedance-2-5-ecommerce-video

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). 运行时通过 scripts/imiva_mcp.py 调用 IMIVA MCP 工具 create_video_generation_task 时会把用户/外部输入的 JSON 参数（如 prompt、素材 HTTPS URL）直接作为 LLM 可读文本/字段传入工具请求；因此若外部用户能提交这些字段，就会形成间接提示注入暴露。

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill invokes npx to run the remote npm package "@infimind/ecom-content-cli@latest" at runtime (see scripts and config), which causes fetching and execution of remote code that the skill relies on to perform MCP calls.