ai-ecommerce-expert-nano-banana-pro-ecommerce-image

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

• Third-party content exposure detected (medium risk: 0.30). Skill 的运行路径通过用户在 create_smart_refine_task 的 prompt/images 参数中提交的自由文本来触发 IMIVA MCP（scripts/imiva_mcp.py 的 tools/call），因此会直接摄入外部用户文本而不需要先选择特定、已固定的条目。

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

• Potentially malicious external URL detected (high risk: 0.90). The skill invokes and runs a remote npm package at runtime via npx ("@infimind/ecom-content-cli@latest"), which causes fetching and executing remote code (scripts/imiva_mcp.py calling npx with PACKAGE), so this external package is a required runtime dependency that can execute code.