Welcome to Zuora Product Documentation

Explore our rich library of product information

Configure Zuora local MCP server

Set up and configure the Zuora local MCP server so MCP-compatible AI clients can securely connect to and interact with your Zuora tenant from your computer.

The Zuora local MCP server enables AI clients that support Model Context Protocol, such as Claude Desktop, Cursor, Windsurf, Claude Code, and OpenAI Codex, to connect to your Zuora tenant from your computer.

MCP is a standard protocol that allows AI assistants to interact with external tools and data sources. With the Zuora local MCP server, your selected AI client can access supported Zuora capabilities through a local Node.js library that runs on your desktop.

Use this guide to set up the Zuora local MCP server on your computer and configure it with your preferred MCP-compatible client. After completing the setup, you can interact with your Zuora tenant from that client.

  1. Enable Zuora AI in the command center. For more detailed information, see Enable Zuora AI.
  2. Create OAuth Credentials to create a MCP client. For more detailed information, see Create OAuth Credentials for local MCP client.
  3. Identify Your MCP Server Endpoint. For more detailed information, see Identify your MCP server endpoint.
  4. If not already installed, install node and open a new terminal window.
  5. Install the Zuora MCP package using npm i zuora-mcp. For more detailed information, see Zuora MCP .
  6. Configure your MCP client.
    Note:

    The installation steps provided in this section cover the most commonly used MCP clients. The Zuora MCP Server follows the standard MCP specification and can be used with any MCP-compatible application. If you want to use the Zuora MCP Server with an application that is not covered in this guide, refer to that application's documentation for installation and configuration instructions.

  7. Validate or test your Zuora MCP server connection. For more detailed information, see Test your Zuora MCP Connection.