PostgreSQL

What is PostgreSQL?

A Model Context Protocol server that provides read-only access to PostgreSQL databases, enabling LLMs to inspect database schemas and execute read-only SQL queries.

How to use PostgreSQL?

Use the server with Claude Desktop by adding configuration to claude_desktop_config.json. Run via Docker (docker run -i --rm mcp/postgres postgresql://...) or NPX (npx -y @modelcontextprotocol/server-postgres postgresql://...). Replace the connection string with your database URL.

Key features of PostgreSQL

• Execute read-only SQL queries via the query tool.
• All queries run within a READ ONLY transaction.
• Automatically discover table schemas as JSON resources.
• Schema resources include column names and data types.
• Connect to any PostgreSQL database via connection string.

Use cases of PostgreSQL

• Let an LLM explore a database schema before generating queries.
• Allow an LLM to run safe, read-only analytical queries.
• Provide structured schema context to improve LLM query accuracy.

FAQ from PostgreSQL

What exactly does this MCP server do?

It provides read-only database access: it can inspect schemas and execute read-only SQL queries, but never writes data.

How do I connect to my PostgreSQL database?

Provide a PostgreSQL connection URL (e.g., postgresql://user:password@host:port/db-name) as a command argument when using Docker or NPX.

Is the server read-only?

Yes. All queries are executed within a READ ONLY transaction, preventing any modifications.

What are the runtime requirements?

The server can be run via Docker (using the mcp/postgres image) or Node.js via NPX (using the @modelcontextprotocol/server-postgres package). No other dependencies are listed.

How is authentication handled?

Authentication is managed through the PostgreSQL connection URL, which can include username and password directly in the URL string.