autoresearch-agent

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

• Third-party content exposure detected (high risk: 0.90). The README's "使用步驟" and rules explicitly require the agent to perform iterative searches and "擷取並結構化每個相關來源的發現（主張、證據、URL、日期、可信度）", which means it will fetch and interpret untrusted public web sources (URLs) as part of its workflow and those findings can materially influence outputs.