multi-agent-orchestration

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

• Third-party content exposure detected (medium risk: 0.65). The skill describes an orchestrator-worker pattern where workers may “fetch web content” (public web pages) and then return summaries to the orchestrator; that runtime ingested outsider text can be embedded into the orchestrator’s LLM context via the worker’s summary/tool results (untrusted prompt-injection risk).