AnythingMCP
Back to Guides

How to Connect PostgreSQL to MCP for AI Agents

Connect PostgreSQL databases directly to AI agents via MCP. Query tables and run reports through natural language with AnythingMCP's read-only database connector.

PostgreSQL: The AI-Ready Database

PostgreSQL is the world's most advanced open-source database. With AnythingMCP's database connector, you can expose your PostgreSQL data directly as MCP tools — letting AI agents interact with your data using natural language.

Important: Read-Only Access

AnythingMCP's database connector is read-only by design. It supports only SELECT queries, with a maximum of 1000 rows per query. This ensures AI agents cannot accidentally modify or delete data.

Auto-Generated Tools

When you create a PostgreSQL database connector, AnythingMCP automatically generates three tools:

| Tool | Description | |---|---| | get_database_schema | Returns all tables, columns, and types | | get_example_queries | Suggests useful queries based on the schema | | execute_query | Executes a read-only SELECT query (max 1000 rows) |

Step-by-Step Guide

Step 1: Deploy AnythingMCP

git clone https://github.com/HelpCode-ai/anythingmcp.git
cd anythingmcp && docker compose up -d

Step 2: Create a Database Connector

Open the AnythingMCP dashboard at http://localhost:3000 and create a new Database connector. Select PostgreSQL as the database type.

Step 3: Configure Connection

Enter your PostgreSQL host, port, database name, username, and password. Always use SSL for production connections.

Step 4: Test the Connection

Use the dashboard to verify the connection and review the auto-discovered schema. The AI agent will use get_database_schema to understand your data structure.

Step 5: Connect to AI Agents

{
  "mcpServers": {
    "postgres": {
      "url": "http://localhost:4000/mcp"
    }
  }
}

AI Agent Use Cases

  • "What were our top 10 products by revenue last month?"
  • "How many new users signed up this week?"
  • "Show me all orders from Germany with total over 200 EUR"
  • "What's the lifetime value of customer john@example.com?"
  • "Search for products containing 'wireless headphones'"

Security Best Practices

  1. Dedicated DB user — Create a PostgreSQL role with SELECT-only permissions
  2. SSL connections — Always use sslmode=require or verify-full
  3. Restrict schemas — Only expose schemas that are safe for AI access
  4. Connection limits — Set max_connections on the PostgreSQL role
  5. Network restrictions — Limit PostgreSQL access to the AnythingMCP server IP

Next Steps