demand-received

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

• Third-party content exposure detected (high risk: 0.80). The skill's workflow explicitly ingests untrusted inputs (user-provided incoming letters) and includes an explicit option to perform and incorporate web/network searches ("搜索网络" / results labeled [联网检索——需复核] and [yuandian检索]) as part of its decision-making and triage outputs (see the "不沉默补充" and "来源标注" sections), so it will read and act on open/public third‑party content that could carry indirect prompt injections.