Connects to Kuzu graph databases to run Cypher queries and inspect schemas. Note: This server is archived as Kuzu database development was discontinued in October 2025.

Provides direct access to Kuzu graph databases through Cypher query execution, schema introspection, and database management with multi-agent coordination, automatic connection recovery, and built-in templates for movies, social networks, and financial data.

8344 views5Local (stdio)

What it does

  • Execute Cypher queries on Kuzu graph databases
  • Inspect database schemas and structure
  • Manage database connections with auto-recovery
  • Generate Cypher queries from natural language
  • Handle multi-agent database access
  • Backup and restore databases via web interface

Best for

Graph data analysis and explorationBuilding AI applications with graph databasesMulti-agent systems needing shared database access
Built-in web management interfaceArchived project - Kuzu development discontinuedMulti-agent coordination support

About KuzuDB

KuzuDB is a community-built MCP server published by jordanburke that provides AI assistants with tools and capabilities via the Model Context Protocol. KuzuDB delivers robust database management with direct access to graph databases, Cypher query execution, and multi-agen It is categorized under databases, analytics data.

How to install

You can install KuzuDB in your AI client of choice. Use the install panel on this page to get one-click setup for Cursor, Claude Desktop, VS Code, and other MCP-compatible clients. This server runs locally on your machine via the stdio transport.

License

KuzuDB is released under the MIT license. This is a permissive open-source license, meaning you can freely use, modify, and distribute the software.

kuzudb-mcp-server

⚠️ ARCHIVED: This project is archived as the Kuzu database repository was archived on October 10, 2025. See ARCHIVE_NOTICE.md for details and alternatives.


A Model Context Protocol server that provides access to Kuzu graph databases. This server enables LLMs to inspect database schemas and execute queries with robust connection recovery, multi-agent coordination, and a built-in web interface.

Archive Status

Archived - October 21, 2025

The Kuzu graph database repository was archived by its maintainers on October 10, 2025 and is now read-only. As Kuzu is no longer actively maintained, this MCP server is also being archived. The project remains fully functional with Kuzu v1.4.1-r.4. See ARCHIVE_NOTICE.md for full details, technical achievements, and alternative graph database options.

🚀 Key Features

  • 📊 Web UI: Built-in database management interface with backup/restore capabilities
  • 🔐 Authentication: OAuth and Basic Auth support for secure access
  • 🤝 Multi-Agent: Safe concurrent access from multiple AI agents (experimental)
  • 🔄 Auto-Recovery: Automatic connection recovery with exponential backoff
  • 🐳 Docker Ready: Pre-built images and docker-compose workflow
  • 📱 Dual Transport: Both stdio and HTTP transport modes
  • 🧠 AI-Powered: Natural language to Cypher query generation

Quick Start

Install and Test

# Install globally
npm install -g kuzudb-mcp-server

# Quick test with auto-created database
pnpm serve:test              # stdio transport (default)
pnpm serve:test:http         # HTTP transport with Web UI
pnpm serve:test:inspect      # HTTP with MCP Inspector

# Server management
pnpm kill    # Stop running servers
pnpm restart # Restart with HTTP transport

Development Setup

# Clone and setup
git clone https://github.com/jordanburke/kuzudb-mcp-server.git
cd kuzudb-mcp-server
pnpm install

# Initialize databases
pnpm db:init                 # Empty test database
pnpm db:init:movies          # Sample movie data

One-Line Docker Setup

# Pull and run with mounted database
docker run -d -p 3000:3000 -p 3001:3001 \
  -v /path/to/your/database:/database \
  ghcr.io/jordanburke/kuzudb-mcp-server:latest

# Access Web UI at http://localhost:3001/admin
# MCP endpoint at http://localhost:3000/mcp

Components

Tools

  • getSchema - Fetch complete database schema (nodes, relationships, properties)
  • query - Execute Cypher queries with automatic error recovery

Prompts

  • generateKuzuCypher - Convert natural language to Kuzu-specific Cypher queries

🖥️ Web UI for Database Management

The server includes a powerful web interface that automatically starts with HTTP transport.

Features

  • 📁 Database Backup & Restore: Download .kuzu backups and restore from browser
  • 📤 Direct File Upload: Upload existing Kuzu database files (main + .wal)
  • 📊 Database Info: View path, mode, connection status, and schema statistics
  • 🔒 Secure Access: Optional authentication protection
  • 👁️ Read-Only Support: Upload/restore disabled in read-only mode

Quick Access

# Start with Web UI (auto-enabled with HTTP)
pnpm serve:test:http

# Access Web UI
open http://localhost:3001/admin

Docker with Web UI

# Using docker-compose (recommended)
docker-compose up -d
open http://localhost:3001/admin

# Manual Docker with Web UI
docker run -d \
  -p 3000:3000 -p 3001:3001 \
  -v /path/to/database:/database \
  -e KUZU_WEB_UI_AUTH_USER=admin \
  -e KUZU_WEB_UI_AUTH_PASSWORD=changeme \
  ghcr.io/jordanburke/kuzudb-mcp-server:latest

API Endpoints

  • /admin - Main web interface
  • /health - Health check endpoint
  • /api/info - Database information (JSON)
  • /api/backup - Download database backup
  • /api/restore - Upload and restore database

🔐 Authentication & Security

The server supports two authentication methods for different use cases:

OAuth (Production Recommended)

Best for production deployments with token-based security:

# Testing OAuth locally
pnpm serve:test:http:oauth     # admin/secret123
pnpm serve:test:inspect:oauth  # With MCP Inspector

# Production OAuth setup
KUZU_OAUTH_ENABLED=true \
KUZU_OAUTH_USERNAME=admin \
KUZU_OAUTH_PASSWORD=your-secure-password \
KUZU_OAUTH_USER_ID=admin-user \
[email protected] \
KUZU_JWT_EXPIRES_IN=31536000 \
node dist/index.js /path/to/database --transport http

Basic Auth (Development/Testing)

Simpler setup for development and testing:

# Testing Basic Auth locally  
pnpm serve:test:http:basic     # admin/secret123
pnpm serve:test:inspect:basic  # With MCP Inspector

# Production Basic Auth setup
KUZU_BASIC_AUTH_USERNAME=admin \
KUZU_BASIC_AUTH_PASSWORD=your-secure-password \
KUZU_BASIC_AUTH_USER_ID=admin-user \
[email protected] \
node dist/index.js /path/to/database --transport http

Web UI Authentication

Secure the Web UI interface:

# Add Web UI authentication
KUZU_WEB_UI_AUTH_USER=admin \
KUZU_WEB_UI_AUTH_PASSWORD=changeme \
node dist/index.js /path/to/database --transport http

JWT Token Configuration

Configure JWT token lifetime (OAuth mode only):

# Set token expiration in seconds (default: 31536000 = 1 year)
KUZU_JWT_EXPIRES_IN=3600    # 1 hour
KUZU_JWT_EXPIRES_IN=86400   # 24 hours
KUZU_JWT_EXPIRES_IN=2592000 # 30 days

Security Recommendations

  • Always use authentication for production deployments
  • Use OAuth for external-facing servers
  • Use Basic Auth for internal development/testing
  • Enable Web UI auth when exposing the interface
  • Use HTTPS in production environments
  • Configure JWT expiration based on your security requirements

Usage with Claude Desktop

Docker (Recommended)

{
  "mcpServers": {
    "kuzu": {
      "command": "docker",
      "args": [
        "run", "-v", "/path/to/database:/database",
        "--rm", "-i", "ghcr.io/jordanburke/kuzudb-mcp-server:latest"
      ]
    }
  }
}

npm/npx

{
  "mcpServers": {
    "kuzu": {
      "command": "npx",
      "args": ["kuzudb-mcp-server", "/path/to/database"]
    }
  }
}

Smithery (Easiest)

# Install via Smithery - includes sample database
smithery install kuzudb-mcp-server

Environment Variables

{
  "mcpServers": {
    "kuzu": {
      "command": "npx",
      "args": ["kuzudb-mcp-server"],
      "env": {
        "KUZU_MCP_DATABASE_PATH": "/path/to/database",
        "KUZU_READ_ONLY": "true"
      }
    }
  }
}

🌐 Remote Connection (HTTP Transport)

Pre-built Docker Images

# Pull latest image
docker pull ghcr.io/jordanburke/kuzudb-mcp-server:latest

# Run with custom configuration
docker run -d \
  -p 3000:3000 -p 3001:3001 \
  -v /path/to/database:/database \
  -e KUZU_READ_ONLY=false \
  ghcr.io/jordanburke/kuzudb-mcp-server:latest

Local Development

# HTTP server mode
node dist/index.js /path/to/database --transport http --port 3000

# With custom endpoint
node dist/index.js /path/to/database --transport http --port 8080 --endpoint /kuzu

MCP Inspector Testing

# Auto-start inspector
pnpm serve:test:inspect

# Manual setup
node dist/index.js /path/to/database --transport http
npx @modelcontextprotocol/inspector http://localhost:3000/mcp

Remote Client Configuration

{
  "mcpServers": {
    "kuzu-remote": {
      "uri": "http://localhost:3000/mcp",
      "transport": "http"
    }
  }
}

🤝 Multi-Agent Coordination (Experimental)

Enable safe concurrent access from multiple AI agents (e.g., Claude Desktop + Claude Code):

Configuration

{
  "mcpServers": {
    "kuzu": {
      "command": "npx",
      "args": ["kuzudb-mcp-server", "/path/to/database"],
      "env": {
        "KUZU_MULTI_AGENT": "true",
        "KUZU_AGENT_ID": "claude-desktop",
        "KUZU_LOCK_TIMEOUT": "10000"
      }
    }
  }
}

How It Works

  • Read queries: Execute immediately without coordination
  • Write queries: Acquire exclusive file-based locks
  • Auto cleanup: Stale locks detected and removed
  • Clear errors: Lock conflicts return helpful retry messages

Important Notes

  • Experimental feature for local development
  • Both agents must use the same database path
  • Lock files created in database directory
  • 10-second default timeout covers most operations

🛠️ Development

Build and Test

# Install dependencies
pnpm install

# Build project
pnpm build

# Development with watch
pnpm dev

# Run tests
pnpm test
pnpm test:ui
pnpm test:coverage

# Linting and formatting
pnpm lint
pnpm typecheck
pnpm format:check

Local Claude Desktop Setup

{
  "mcpServers": {
    "kuzu": {
      "command": "node",
      "args": [
        "/path/to/kuzudb-mcp-server/dist/index.js",
        "/path/to/database"
      ]
    }
  }
}

🔧 Environment Variables Reference

VariableDescriptionDefaultUsage
Database
KUZU_MCP_DATABASE_PATHDatabase path if not in args-Startup
KUZU_READ_ONLYEnable read-only modefalseSecurity
Connection
KUZU_MAX_RETRIESConnection recovery attempts2Reliability
Multi-Agent
KUZU_MULTI_AGENTEnable coordinationfalseConcurrency
KUZU_AGENT_IDUnique agent identifierunknown-{pid}Locking
KUZU_LOCK_TIMEOUTLock timeout (ms)10000Performance
Web UI
KUZU_WEB_UI_ENABLEDEnable/disable Web UItrueInterface
KUZU_WEB_UI_PORTWeb UI port3001Network
KUZU_WEB_UI_AUTH_USERWeb UI username-Security
KUZU_WEB_UI_AUTH_PASSWORDWeb UI password-Security
Authentication
KUZU_OAUTH_ENABLEDEnable OAuthfalseSecurity
`KUZU_OAUTH_

README truncated. View full README on GitHub.

Alternatives

Related Skills

Browse all skills
literature-review

Conduct comprehensive, systematic literature reviews using multiple academic databases (PubMed, arXiv, bioRxiv, Semantic Scholar, etc.). This skill should be used when conducting systematic literature reviews, meta-analyses, research synthesis, or comprehensive literature searches across biomedical, scientific, and technical domains. Creates professionally formatted markdown documents and PDFs with verified citations in multiple citation styles (APA, Nature, Vancouver, etc.).

377
postgresql-psql

Comprehensive guide for PostgreSQL psql - the interactive terminal client for PostgreSQL. Use when connecting to PostgreSQL databases, executing queries, managing databases/tables, configuring connection options, formatting output, writing scripts, managing transactions, and using advanced psql features for database administration and development.

38
data-storytelling

Transform data into compelling narratives using visualization, context, and persuasive structure. Use when presenting analytics to stakeholders, creating data reports, or building executive presentations.

27
content-trend-researcher

Advanced content and topic research skill that analyzes trends across Google Analytics, Google Trends, Substack, Medium, Reddit, LinkedIn, X, blogs, podcasts, and YouTube to generate data-driven article outlines based on user intent analysis

23
data-scientist

Expert data scientist for advanced analytics, machine learning, and statistical modeling. Handles complex data analysis, predictive modeling, and business intelligence. Use PROACTIVELY for data analysis tasks, ML modeling, statistical analysis, and data-driven insights.

13
google-analytics

Analyze Google Analytics data, review website performance metrics, identify traffic patterns, and suggest data-driven improvements. Use when the user asks about analytics, website metrics, traffic analysis, conversion rates, user behavior, or performance optimization.

13