research

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

• Third-party content exposure detected (high risk: 1.00). This skill launches web research via run_research_task / get_research_task_result that gathers and synthesizes content from open web sources (news, academic sites, social media, public websites and arbitrary source_urls), so the agent will read untrusted, user-generated third‑party content as part of its workflow.