Your n8n Agent Has Amnesia. Give It a Knowledge Graph

FalkorDB’s build log describes an n8n workflow that turns a GitHub-hosted handbook into a continuously updated GraphRAG knowledge base accessible through a chat interface. A Chat Trigger handles team questions while a GitHub Trigger detects merged changes to Markdown files, with both paths converging on an AI Agent that uses five GraphRAG tools to ingest repositories or text, query the graph, list documents, and update documents. The hosted GraphRAG service stores document chunks and embeddings alongside extracted entities, relationships, and provenance links in a managed FalkorDB property graph, allowing answers to follow connections across documents and cite sources rather than relying only on similarity search. n8n supplies the workflow orchestration, webhook registration, credential management, model routing, chat memory, and GitHub synchronization, while GraphRAG separately uses configured LLM providers for extraction, retrieval, answer composition, and embeddings. The instructions cover setup of GraphRAG credentials, n8n community nodes, public webhook access, agent rules, and GitHub integration, emphasizing that the agent should query the graph for factual questions and use complete document contents for updates. To reduce update costs, the system hashes documents and chunks, reuses unchanged chunks and their graph relationships, processes only modified content with LLM extraction, and atomically replaces stale graph data after updates.

Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.