NeuronDB: AI Database Extension for PostgreSQL

Model Context Protocol server enabling MCP-compatible clients (like Claude Desktop) to access NeuronDB through stdio communication. JSON-RPC 2.0 implementation.

NeuronDB A PostgreSQL AI-Extension Demo

NeuronMCP Protocol Demo
NeuronMCP Protocol Demo Terminal
Interactive demonstration of NeuronMCP JSON-RPC 2.0 protocol over stdio for MCP-compatible clients.
JSON-RPC 2.0stdio TransportMCP ProtocolClaude Desktop
Click "Run Demo" to see NeuronMCP protocol in action

Architecture

MCP server architecture with protocol handling, tools, resources, and middleware

NeuronMCP Architecture

MCP Client

Claude Desktop | Custom Clients | stdio Communication

Protocol Handler

  • • JSON-RPC 2.0
  • • stdio Transport
  • • Request Routing
  • • Error Handling

Tools

  • • Vector Operations
  • • ML Operations
  • • Analytics
  • • RAG Operations

Resources

  • • Schema Info
  • • Model Catalog
  • • Index Config
  • • Stats & Config

Middleware

Validation
Input checks
Logging
Structured logs
Timeout
Request limits
Error Handling
Graceful errors

NeuronDB PostgreSQL

Vector Search | ML Algorithms | Embeddings | SQL Functions

MCP Server Features

Why NeuronDB

MCP Protocol Implementation

Full JSON-RPC 2.0 implementation with stdio communication protocol. Request routing and error handling. Compatible with all MCP-compatible clients including Claude Desktop. Protocol version negotiation and capability discovery.

Vector Operations Tools

Complete set of tools for vector search, similarity computation, embedding generation, and index creation. Direct access to NeuronDB vector engine. Support for multiple vector types and distance metrics. Batch operations support.

ML Operations Tools

Model training, prediction, and evaluation tools. Access to all 52 ML algorithms in NeuronDB. Model catalog management. Hyperparameter tuning support. Model versioning and A/B testing capabilities.

Analytics Tools

Data analysis, clustering, and dimensionality reduction tools. Quality metrics computation. Outlier detection. Time series analysis. Comprehensive analytics suite accessible via MCP protocol.

RAG Operations

Document processing, context retrieval, and response generation tools. Complete RAG pipeline accessible through MCP. Semantic search with reranking. LLM integration for answer generation.

Resource Management

Schema information, model catalog, index configurations, and system statistics exposed as MCP resources. Real-time resource updates. Resource discovery and metadata access.

Middleware Support

Validation middleware for input sanitization. Structured logging for debugging and monitoring. Request timeout handling. Comprehensive error handling with graceful degradation.

Claude Desktop Integration

Ready-to-use configuration for Claude Desktop. Docker and binary deployment options. Environment-based configuration. Seamless integration with existing NeuronDB installations.

Capabilities

MCP server features

CapabilityDescriptionPerformanceProduction Ready
MCP ProtocolJSON-RPC 2.0 over stdioLow-latency communication✓
Vector ToolsSearch, embeddings, indexingSub-second operations✓
ML ToolsTraining, prediction, evaluationGPU-accelerated✓
Analytics ToolsClustering, analysis, metricsEfficient algorithms✓
RAG ToolsDocument processing, retrievalEnd-to-end optimization✓
ResourcesSchema, models, indexes, statsReal-time updates✓
MiddlewareValidation, logging, timeoutLow overhead✓
Claude DesktopReady-to-use configurationEasy setup✓
Get Started

Connect Claude Desktop to NeuronDB

Deploy NeuronMCP. Enable MCP-compatible clients to access NeuronDB vector search, ML algorithms, and RAG capabilities.

View DocumentationView on GitHub