multi-agent-patterns

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

• Third-party content exposure detected (high risk: 0.90). The skill includes agents that perform web search and browsing (e.g., references/frameworks.md "LangGraph Swarms" creates a research agent with tools [search, browse, read] and researcher_node calls perform_research), which indicates the agent will fetch and read arbitrary public web content and thus can ingest untrusted third-party material.