ai-ecommerce-expert-gpt-image-2-ecommerce-image

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

• Third-party content exposure detected (medium risk: 0.30). 运行时工作流通过用户在 Skill 调用 create_smart_refine_task 时提供的自由文本（如 prompt 字段）直接作为 LLM 任务输入发送给 IMIVA MCP，而该文本由外部用户作者产生且在未先选择特定“可信条目”的情况下被消费。

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 fetch and execute the npm package "@infimind/ecom-content-cli@latest" at runtime (used by scripts/imiva_mcp.py and in the config), which runs remote code and is a required runtime dependency.