cn-blog

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

• Third-party content exposure detected (high risk: 0.75). At runtime the script fetches first-party Zhihu API JSON over the network (e.g., GET https://www.zhihu.com/api/v4/me, .../members/{token}/articles, .../articles/{id}) and then parses/prints the returned text into the agent context; this is outsider-authored free text from public web content.