multi-agent-patterns

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

• Third-party content exposure detected (high risk: 0.90). This skill instructs researcher agents to perform web search/browsing and document retrieval (see researcher_node's perform_research and the swarm create_agent calls that list tools [search, browse, read]), which means it ingests arbitrary public web/user-generated content from the open web.