MCP PostgreSQL Server

MCP PostgreSQL Server is an MCP server that enables AI models to interact with PostgreSQL databases through a standardized interface. Built for developers, data analysts, and software engineers, it connects Model Context Protocol clients directly to relational database instances. The server allows agents to inspect database structure by listing schemas, enumerating tables within specific schemas, and retrieving column definitions and table designs. Beyond schema inspection, it allows AI tools to perform parameterized read queries using SELECT statements, as well as write operations such as INSERT, UPDATE, and DELETE. It features prepared statement support to safeguard against SQL injection, accepting both standard PostgreSQL dollar placeholders and MySQL-style question mark placeholders. The server handles database connections through standard configuration variables, provides automatic connection cleanup, and offers validation and detailed error reporting for missing parameters and invalid queries during execution.

Category: Databases & Data Stores

Tags: database, postgresql, relational, SQL

Visit MCP PostgreSQL Server

How to install and configure MCP PostgreSQL Server

1. Ensure Node.js and npm are installed on your system. 2. Add the server to your MCP client configuration file (such as Claude Desktop or Cursor) within the mcpServers object using the following stdio definition: json { "mcpServers": { "postgres": { "type": "stdio", "command": "npx", "args": ["-y", "mcp-postgres-server"], "env": { "PG_HOST": "your_host", "PG_PORT": "5432", "PG_USER": "your_user", "PG_PASSWORD": "your_password", "PG_DATABASE": "your_database" } } } } 3. Replace the environment variable placeholders (PG_HOST, PG_PORT, PG_USER, PG_PASSWORD, PG_DATABASE) with your actual PostgreSQL database credentials. 4. Save the configuration file and restart your MCP client to initialize the server.

What you can do with MCP PostgreSQL Server

• Discovering database architecture by listing schemas and table definitions across public or custom PostgreSQL schemas. - Executing safe read-only SQL queries with prepared statement parameters to extract analytics and user data. - Inserting new database records or applying batch updates to existing rows through parameterized execution. - Debugging database query issues and inspecting column types directly through conversational AI model interactions. - Establishing ad-hoc connections to alternative PostgreSQL instances using the dynamic connection tool during client runtime.

Key facts

• Open Source
• https://github.com/antonorlov/mcp-postgres-server
• Databases & Data Stores
• database, postgresql, relational, SQL

Part of MCP Servers

How do I install MCP PostgreSQL Server?

You can install MCP PostgreSQL Server via npm using the command npm install mcp-postgres-server or run it directly using npx. To register it with an MCP client, add the npx execution configuration and database environment variables such as PG_HOST, PG_PORT, PG_USER, PG_PASSWORD, and PG_DATABASE into your client configuration file.

What can MCP PostgreSQL Server do?

The server provides tools to connect to PostgreSQL databases, list schemas, list tables within schemas, describe table structures, and run parameterized queries. It supports SELECT operations via the query tool and handles data modifications including INSERT, UPDATE, and DELETE statements via the execute tool.

Does MCP PostgreSQL Server protect against SQL injection?

Yes, MCP PostgreSQL Server uses prepared statements for both read and write operations to protect against SQL injection. It supports parameter bindings using standard PostgreSQL numeric placeholders like $1 and $2, as well as MySQL-style question mark placeholders.

Is MCP PostgreSQL Server open source?

Yes, MCP PostgreSQL Server is an open-source project distributed under the MIT license. You can view its source code, contribute, and report issues through its repository on GitHub.