Setting Up MCP in Cursor: Complete Configuration Guide

What is MCP in Cursor?

Cursor is an AI-first code editor built on VS Code that integrates the Model Context Protocol (MCP) to connect to external tools and data sources. Instead of explaining your project structure repeatedly, you can integrate directly with your tools.

MCP servers expose capabilities through the protocol, connecting Cursor to external tools or data sources. Once configured, Agent automatically uses MCP tools when relevant to your prompts, making your AI assistant significantly more capable.

Transport Types

Cursor supports three transport methods for MCP servers:

Protocol Support

Cursor supports all major MCP protocol capabilities:

One-Click Installation

The easiest way to install MCP servers in Cursor is through the built-in collection:

1. Open Cursor Settings (Cmd+Shift+J)
2. Navigate to Features → Model Context Protocol
3. Browse available MCP servers
4. Click “Add to Cursor” to install
5. Authenticate with OAuth if required

Configuring mcp.json

For custom MCP servers or advanced configuration, use the mcp.json file:

Configuration Locations

Configuration Examples

{
  "mcpServers": {
    "filesystem": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-filesystem", "/path/to/dir"],
      "env": {
        "API_KEY": "your-key"
      }
    }
  }
}

{
  "mcpServers": {
    "my-python-server": {
      "command": "python",
      "args": ["mcp-server.py"],
      "env": {
        "API_KEY": "your-key"
      }
    }
  }
}

{
  "mcpServers": {
    "remote-api": {
      "url": "https://api.example.com/mcp",
      "headers": {
        "Authorization": "Bearer ${env:MY_TOKEN}"
      }
    }
  }
}

STDIO Server Configuration

For local command-line servers, configure these fields in your mcp.json:

Variable Interpolation

Cursor supports variable interpolation in mcp.json values. Variables can be used incommand,args,env,url, andheaders fields:

Example with Variables

{
  "mcpServers": {
    "local-server": {
      "command": "python",
      "args": ["${workspaceFolder}/tools/mcp_server.py"],
      "env": {
        "API_KEY": "${env:API_KEY}",
        "HOME_DIR": "${userHome}"
      }
    },
    "remote-server": {
      "url": "https://api.example.com/mcp",
      "headers": {
        "Authorization": "Bearer ${env:MY_SERVICE_TOKEN}"
      }
    }
  }
}

Using MCP in Chat

Once configured, Agent automatically uses MCP tools listed under “Available Tools” when relevant. This includes Plan Mode.

Toggling Tools

Enable or disable MCP tools directly from the chat interface. Click a tool name in the tools list to toggle it. Disabled tools won't be loaded into context or available to Agent.

Tool Approval

Agent asks for approval before using MCP tools by default. Click the arrow next to the tool name to see arguments before approving.

Auto-Run Mode

Enable auto-run for Agent to use MCP tools without asking. This works similarly to terminal commands and can be configured per-tool or globally in settings.

Security Considerations

When installing MCP servers, follow these security practices:

• ⚠Verify the source: Only install MCP servers from trusted developers and repositories.
• ⚠Review permissions: Check what data and APIs the server will access.
• ⚠Limit API keys: Use restricted API keys with minimal required permissions.
• ⚠Audit code: For critical integrations, review the server's source code.

Troubleshooting

References & Further Reading