writing-assistant

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

• Third-party content exposure detected (high risk: 1.00). Yes — the required "先验调研" step in content/methods/01-选题.md explicitly mandates searching public, user-generated platforms (小红书/知乎/微信搜一搜) and judging result counts/quality, which requires ingesting and interpreting untrusted third‑party content that can directly change topic selection and subsequent actions.