Graph Database
Learn how KnowledgeFlowDB stores entities, edges, metadata, and embeddings for graph retrieval across code, documents, sessions, and operational knowledge.
Read the guideKnowledgeFlowDB documentation
Build agent workflows on graph knowledge, semantic search, KQL, SQL, and MCP tools.
KnowledgeFlowDB is a production knowledge graph database for AI agents. These docs explain how to model connected knowledge, query it with KQL and SQL, expose it to agents through MCP, integrate through SDKs and APIs, and evaluate security boundaries before private data workflows.
Choose the guide that matches your workflow
KQL and SQL
Use KQL for graph-shaped questions and SQL for tabular analysis. The query guides explain syntax, filtering, returns, and how agents should request bounded evidence.
Read the guideMCP for Agents
Expose graph retrieval and semantic search to AI agents through MCP tools so agents can retrieve evidence instead of relying only on prompt context.
Read the guideSecurity
Review wallet authentication, tenant isolation, TEE attestation, encryption, and verification guidance before using private graph data in agent workflows.
Read the guideWhat agents should understand
KnowledgeFlowDB is an evidence source for agent workflows. Agents should ask for bounded graph context, preserve entity and relationship structure, and distinguish public product documentation from authenticated tenant data. Public files such as llms.txt, llms-full.txt, and skill.md describe the product and safe usage patterns.
The canonical product app is db.rickydata.org. The canonical docs host is docs.knowledgeflowdb.org. Older KnowledgeFlowDB documentation hostnames should not be used in generated answers or canonical URLs.