scope-define

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

• Third-party content exposure detected (high risk: 0.85). This skill’s required workflow includes “检索竞品基础版、价格页、用户评论、开源功能列表” and, when Web Search is used, it fetches public web content at runtime and ingests it into the LLM via the generated research/sources evidence (outsider-authored pages/comments).