ai-ecommerce-expert-reference-video-generation

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

• Third-party content exposure detected (medium risk: 0.30). 该 Skill 的运行时通过 scripts/imiva_mcp.py 调用 IMIVA MCP 工具 create_video_generation_task，并把用户在 prompt/referenceVideos 等字段中提交的自由文本直接作为工具参数发往 https://imiva.ecpro.com，属于“用户先选/后填”但不需要先选择特定条目来源的间接提示注入风险面。

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

• Potentially malicious external URL detected (high risk: 1.00). The skill invokes npx to fetch and run the npm package "@infimind/ecom-content-cli@latest" at runtime (e.g., via scripts/imiva_mcp.py), which downloads and executes remote code.

MEDIUM W009: Direct money access capability detected (payment gateways, crypto, banking).

• Direct money access detected (high risk: 1.00). The skill exposes explicit MCP APIs that query and consume platform credits and can create billable tasks. It documents and demonstrates get_user_credits and create_video_generation_task (with dryRun, idempotencyKey, and maxCredits) and states that submitting tasks may incur charges. Those calls allow an agent to trigger paid operations that spend user/platform credits, which is direct financial execution for the IMIVA credits system.