hive-create

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

• Third-party content exposure detected (high risk: 1.00). The example agent's research_node explicitly calls web_search and web_scrape (see examples/deep_research_agent/nodes/init.py and the mcp_servers.json declaring hive-tools provides web_search/web_scrape), causing the agent to fetch and analyze arbitrary public web content which it then uses to make workflow decisions (findings, citations, edges), so untrusted third-party content can materially influence behavior.