pgEdge Postgres MCP Server and Natural Language Agent

- Introduction
- Choosing the Right Solution
- Installing the MCP Server
- Deploying on Docker
- Deploying from Source
- Accessing Online Help
- Configuring the MCP Server
- Specifying Configuration Preferences
- Using Environment Variables to Specify Options
- Including Provider Embeddings in a Configuration File
- Configuring the Agent for Multiple Databases
- Configuring Supporting Services; HTTP, systemd, and nginx
- Using an Encryption Secret File
- Enabling or Disabling Features
- Configuring and Using Client Applications
- Connecting with the Web Client
- Using the Go Chat Client
- Using the Claude Desktop
- Authentication and Security
- Authentication Overview
- User Authentication Management
- Token Authentication Management
- Security Best Practices - Checklist
- Security Management
- Reference
- Tools
- Resources
- Prompts
- Examples
- Advanced Topics
- Custom Definitions
- Knowledgebase
- LLM Proxy
- For Developers
- Overview
- MCP Protocol
- API Reference
- Building Chat Clients
- Overview
- Python - Stdio and Claude
- Python - HTTP and OLLAMA
- Contributing
- Development Setup
- Architecture
- Internal Architecture
- KB Builder
- Testing
- CI/CD
- Troubleshooting
- pgEdge Postgres MCP Server and Natural Language Agent Release Notes
- Licence

The pgEdge Postgres Model Context Protocol (MCP) server enables SQL queries against
PostgreSQL databases through MCP-compatible clients like Claude Desktop. The Natural Language Agent provides supporting functionality that allows you to use natural language to form SQL queries.

> 🚧 WARNING: This code is in pre-release status and MUST NOT be put
> into production without thorough testing!

> ⚠️ NOT FOR PUBLIC-FACING APPLICATIONS: This MCP server provides LLMs
> with read access to your entire database schema and data. It should only be
> used for internal tools, developer workflows, or environments where all users
> are trusted. For public-facing applications, consider the
> pgEdge RAG Server instead.
> See the Choosing the Right Solution guide for
> details.

Key Features

- 🔒 Read-Only Protection - All queries run in read-only transactions
- 📊 Resources - Access PostgreSQL statistics and more
- 🛠️ Tools - Query execution, schema analysis, advanced hybrid search
(BM25+MMR), embedding generation, resource reading, and more
- 🧠 Prompts - Guided workflows for semantic search setup, database
exploration, query diagnostics, and more
- 💬 Production Chat Client - Full-featured Go client with Anthropic
prompt caching (90% cost reduction)
- 🌐 HTTP/HTTPS Mode - Direct API access with token authentication
- 🖥️ Web Interface - Modern React-based UI with AI-powered chat for
natural language database interaction
- 🐳 Docker Support - Complete containerized deployment with Docker
Compose
- 🔐 Secure - TLS support, token auth, read-only enforcement
- 🔄 Hot Reload - Automatic reload of authentication files without server
restart

Quick Start

1. Installation

git clone <repository-url>
cd pgedge-postgres-mcp
make build

2. Configure for Claude Code and/or Claude Desktop

Claude Code: .mcp.json in each of your project directories
Claude Desktop on macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
Claude Desktop on Windows: %APPDATA%\\Claude\\claude_desktop_config.json

{
  "mcpServers": {
    "pgedge": {
      "command": "/absolute/path/to/bin/pgedge-postgres-mcp"
    }
  }
}

3. Connect to Your Database

Update your Claude Code and/or Claude Desktop configuration to include database connection
parameters:

{
  "mcpServers": {
    "pgedge": {
      "command": "/absolute/path/to/bin/pgedge-postgres-mcp",
      "env": {
        "PGHOST": "localhost",
        "PGPORT": "5432",
        "PGDATABASE": "mydb",
        "PGUSER": "myuser",
        "PGPASSWORD": "mypass"
      }
    }
  }
}

Alternatively, use a .pgpass file for password management (recommended for
security):

```bash