content-research-writer

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

• Third-party content exposure detected (high risk: 0.70). The required workflow includes “Research [specific topic] and add citations,” which at runtime would typically involve fetching/extracting public web content or other third-party sources (outsider-authored free text) and then inserting that extracted prose into the agent’s LLM context for summarization/citation.