multi-agent-patterns

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

• Third-party content exposure detected (high risk: 0.80). The skill explicitly describes a "Researcher (web search, document retrieval)" sub-agent and discusses searching "multiple independent sources" and links to external resources (e.g., LangGraph docs, arXiv), so it expects agents to fetch and read open/public web content that may be untrusted.