policy-sensitivity-brief

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

• Third-party content exposure detected (high risk: 0.80). The skill explicitly instructs running shared data-fetch scripts and consuming policy/market data from public sources (see references/data-queries.md) and lists "政策数据：政府网站、新闻、AKShare" in references/methodology.md, so the agent ingests open third-party content that can influence analysis and actions.