PostgreSQL MCP Server

中文 | English

A Model Context Protocol (MCP) server that provides tools for interacting with a PostgreSQL database. It enables AI assistants to execute SQL queries, explain statements, create tables, and list database tables via the MCP protocol.

✨ Features

• Interact with Databases via AI: Enables LLMs to perform database operations through a structured protocol.
• Secure Toolset: Separates read and write operations into distinct, authorizable tools (read_query, write_query).
• Schema Management: Allows for table creation (create_table) and listing (list_tables).
• Query Analysis: Provides a tool to analyze query execution plans (explain_query).
• Multiple Transport Modes: Supports stdio, Server-Sent Events (sse), and streamableHttp for flexible client integration.
• Environment-Based Configuration: Easily configurable using a .env file.

🛠️ Available Tools

The server exposes the following tools for MCP clients to invoke:

Tool Name	Description	Parameters
read_query	Executes a SELECT SQL query.	query (string, required): The SELECT statement to execute.
write_query	Executes an INSERT, UPDATE, or DELETE SQL query.	query (string, required): The INSERT/UPDATE/DELETE statement to execute.
create_table	Executes a CREATE TABLE SQL statement.	schema (string, required): The CREATE TABLE statement.
list_tables	Lists all user-created tables in the database.	schema (string, optional): The schema name to filter tables by.
explain_query	Returns the execution plan for a given SQL query.	query (string, required): The query to explain (must start with EXPLAIN).

🚀 Quick Start

Prerequisites

• Go 1.23 or later
• A PostgreSQL database server

Installation

1. Clone the repository:

   git clone https://github.com/leixiaotian1/pgsql-mcp-server.git
   cd pgsql-mcp-server
   

2. Install dependencies:

   go mod download
   

3. Build the MCP server:

   go build -o pgsql-mcp-server
   

Configuration

The pg-mcp-server requires database connection details to be provided via environment variables. Create a .env file in the project root with the following variables:

DB_HOST=localhost      # PostgreSQL server host
DB_PORT=5432           # PostgreSQL server port
DB_NAME=postgres       # Database name
DB_USER=your_username  # Database user
DB_PASSWORD=your_pass  # Database password
DB_SSLMODE=disable     # SSL mode (disable, require, verify-ca, verify-full)
SERVER_MODE=stdio      # Server mode (stdio, sse, streamableHttp)

Usage

Running the Server

./pgsql-mcp-server

MCP Configuration

To use this server with an MCP-enabled AI assistant, add the following to your MCP configuration:

{
  "mcpServers": {
    "pgsql-mcp-server": {
      "command": "/path/to/pgsql-mcp-server",
      "args": [],
      "env": {
        "DB_HOST": "localhost",
        "DB_PORT": "5432",
        "DB_NAME": "postgres",
        "DB_USER": "your_username",
        "DB_PASSWORD": "your_password",
        "DB_SSLMODE": "disable",
        "SERVER_MODE": "stdio"
      },
      "disabled": false,
      "autoApprove": []
    }
  }
}

---

DOCKER DEPLOYMENT

Prerequisites

• Docker installed

Deployment Steps

1. Clone the project

   git clone https://github.com/leixiaotian1/pgsql-mcp-server.git
   cd pgsql-mcp-server
   

2. Configure .env file

   Create a .env file in the project root directory. This file stores database connection information. Ensure the DB_HOST value matches the database container name you'll start later.

   DB_HOST=postgres
   DB_PORT=5432
   DB_NAME=postgres
   DB_USER=user
   DB_PASSWORD=password
   DB_SSLMODE=disable
   SERVER_MODE=sse
   

3. Create Docker network

   To enable communication between the application container and database container, create a shared Docker network. This command only needs to run once.

   docker network create sql-mcp-network
   

4. Start PostgreSQL database container

   Use this command to start a PostgreSQL container and connect it to our network.

   Note:

   • --name postgres: Container name, must exactly match the DB_HOST in your .env file.
   • --network sql-mcp-network: Connect to the shared network.
   • -p 5432:5432: Maps host's 5432 port to container's 5432 port. This means you can connect from your computer (e.g., using DBeaver) via localhost:5432, while the app container will access 5432 port directly through the internal network.

   docker run -d \
     --name postgres \
     --network sql-mcp-network \
     -e POSTGRES_USER=user \
     -e POSTGRES_PASSWORD=password \
     -e POSTGRES_DB=postgres \
     -p 5432:5432 \
     postgres
   

5. Build and run the application

   Now you can use commands from the Makefile to manage the application.

   • Build image and run container:

     make build
     make run
     

     This will automatically stop old containers, build a new image, and start a new container.

   • View application logs:

     make logs
     

     If you see Successfully connected to database, everything is working correctly.

   • Stop the application:

     make stop
     

---

🔌 Server Modes

You can select the transport protocol by setting the SERVER_MODE environment variable.

stdio

The server communicates over standard input and output. This is the default mode and is ideal for local testing or direct integration with command-line-based MCP clients.

sse

The server communicates using Server-Sent Events (SSE). When this mode is enabled, the server will start an HTTP service and listen for connections.

• SSE Endpoint: http://localhost:8088/sse
• Message Endpoint: http://localhost:8088/message

streamableHttp

The server uses the Streamable HTTP transport, a more modern and flexible HTTP-based transport for MCP.

• Endpoint: http://localhost:8088/mcp

🤝 Contributing

Contributions are welcome! If you find any bugs, have feature requests, or suggestions for improvement, please feel free to submit a Pull Request or open an Issue.

1. Fork the Project.
2. Create your Feature Branch (git checkout -b feature/AmazingFeature).
3. Commit your Changes (git commit -m 'Add some AmazingFeature').
4. Push to the Branch (git push origin feature/AmazingFeature).
5. Open a Pull Request.

📄 License

This project is open source and is licensed under the MIT License.