Postgres MCP Server

🚀 PostgreSQL MCP Server (Model Context Protocol)

A basic implementation of the Model Context Protocol (MCP) for direct interaction with PostgreSQL databases.

This MCP enables you to transform your AI chat experience in the following ways:

• Store and retrieve chat data in a PostgreSQL database.
• Create AI applications using persistent data storage.
• Build knowledge management systems based on structured database queries.
• Analyze large datasets directly through natural language prompts.
• Implement database - based workflows without relying on traditional coding.

This implementation uses the MCP protocol to securely connect Claude AI to a PostgreSQL database, allowing Claude to perform operations on your behalf while maintaining appropriate isolation and security boundaries.

✨ Features

• Execute queries on a PostgreSQL database.
• Manage tables (create, delete).
• Perform data operations (select, insert, update, delete).
• Check database schemas.
• Integrate with Claude via the MCP protocol.

📦 Prerequisites

• Python 3.8+
• A PostgreSQL server
• Access to Claude AI with MCP capabilities

📦 Installation

1. Clone this repository to your local machine.
2. Create and activate a Python virtual environment:

# Create a virtual environment
python -m venv .mcp

# Activate the virtual environment
# On macOS/Linux
source .mcp/bin/activate
# On Windows
.mcp\Scripts\activate

3. Install the required dependencies:

pip install -r requirements.txt

📚 Configuration

1. Create a .env file in the project root directory, containing your PostgreSQL connection details and debugging information.
2. Configure the PostgreSQL MCP server with Claude by adding the following configuration to the MCP configuration section of the Claude AI application settings:

{
  "mcpServers": {
    "PostgreSQL MCP": {
      "command": "<actual_path>/.mcp/bin/uv",
      "args": [
        "run",
        "--with",
        "mcp[cli]",
        "--with",
        "asyncpg",
        "--with",
        "httpx",
        "--with",
        "python-dotenv",
        "--with",
        "psycopg2-binary",
        "mcp",
        "run",
        "<actual_path>/postgres_mcp_server.py"
      ],
      "env": {
        "POSTGRES_HOST": "<your_PostgreSQL_host>",
        "POSTGRES_PORT": "<your_PostgreSQL_port>",
        "POSTGRES_USER": "<your_username>",
        "POSTGRES_PASSWORD": "<your_password>",
        "POSTGRES_DB": "<your_database_name>"
      }
    }
  }
}

Note: Replace <actual_path> with the actual path.

Add this configuration to the MCP configuration section of the Claude AI application settings. Then, Claude will be able to connect to your PostgreSQL MCP server.