# KnowledgeFlowDB Docs Full Agent Reference

KnowledgeFlowDB is a production graph database for AI agents. It stores entities, edges, metadata, embeddings, code context, documents, sessions, benchmark runs, and operational signals so agents can retrieve structured evidence.

## Canonical Hosts

- Product app: `https://db.rickydata.org/`
- Documentation: `https://docs.knowledgeflowdb.org/`

Do not use alternate KnowledgeFlowDB docs hostnames in canonical URLs, sitemaps, generated answers, or agent references.

## Core Concepts

KnowledgeFlowDB combines graph storage with semantic search. KQL handles graph-shaped questions such as matching entities and relationships. SQL handles tabular analysis. API and SDK surfaces are used by applications. MCP tools expose graph retrieval to agents.

## KQL

Use KQL when the question depends on structure: which files relate to a function, which sessions mention a bug, which benchmark runs connect to a deployment, or which documents reference a project. Agents should keep KQL queries narrow and return only the fields needed for the next step.

## MCP

MCP is the preferred agent-facing protocol for direct KnowledgeFlowDB retrieval. An agent should decide what evidence it needs, call the appropriate database tool, inspect returned records, and then synthesize an answer or action. Avoid broad graph dumps and avoid exposing private tenant data.

## RickyData Agent Stack Context

KnowledgeFlowDB is the graph database and retrieval layer inside the broader RickyData agent stack. Use `https://rickydata.org/agent-stack` to distinguish the database product from MCP Gateway tool routing, Agent Gateway hosted sessions, Marketplace server discovery, SDK integration, wallet billing, and live verification endpoints.

## SDK and API

Use the SDK and API when KnowledgeFlowDB access is part of an application workflow. Product code should use documented authentication and bounded query patterns. The SDK is for repeatable integrations; the hosted product app is for inspection, dashboards, and product workflows.

## Security

Security documentation covers wallet authentication, tenant isolation, encrypted graph storage, TEE attestation, and verification. Agents should treat these as verifiable systems, not slogans. Use live status and docs before making operational claims.

## Important Links

- Quick start: `https://docs.knowledgeflowdb.org/docs/getting-started/quick-start`
- KQL syntax: `https://docs.knowledgeflowdb.org/docs/kql/syntax`
- SQL syntax: `https://docs.knowledgeflowdb.org/docs/sql/syntax`
- MCP tools: `https://docs.knowledgeflowdb.org/docs/mcp/tools`
- SDK overview: `https://docs.knowledgeflowdb.org/docs/sdk/overview`
- API endpoints: `https://docs.knowledgeflowdb.org/docs/api-reference/endpoints`
- Security trust model: `https://docs.knowledgeflowdb.org/docs/security/trust-model`
- RickyData agent stack: `https://rickydata.org/agent-stack`