research-synthesize

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

• Third-party content exposure detected (medium risk: 0.65). The required workflow ingests “raw findings” and “references” from the mcp__claude-flow__memory_search namespaces (research and research-sources), which can include outsider-authored free text that was previously fetched/collected during deep-research, and then synthesizes it via mcp__claude-flow__agentdb_context-synthesize into the LLM context.