ai-ecommerce-expert-taobao-ecommerce-content

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

• Third-party content exposure detected (medium risk: 0.30). SKILL.md 的运行时工作流通过 IMIVA MCP 工具调用（如 create_product_main_image_task/get_user_tasks）接收用户在聊天/自然语言中提供的“商品卖点、规格、使用场景、文案”等自由文本并将其纳入工具参数，属于用户可随意提交的间接提示注入输入链路。

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

• Potentially malicious external URL detected (high risk: 1.00). The skill's runtime invokes npx to fetch-and-run the npm package "@infimind/ecom-content-cli@latest" (e.g., in scripts/imiva_mcp.py and SKILL.md), which causes remote code to be downloaded and executed at runtime, so "@infimind/ecom-content-cli@latest" is flagged.