Setting Up MCP Servers

Where MCPs are configured depends on which AI tool you’re using.

This guide covers setup for the most popular platforms.

Quick Reference

Claude Desktop Setup

Step 1: Find the Config File

Mac:

~/Library/Application Support/Claude/claude_desktop_config.json

Windows:

%APPDATA%\Claude\claude_desktop_config.json

Step 2: Add MCP Server

Open the file and add:

{
  "mcpServers": {
    "filesystem": {
      "command": "npx",
      "args": [
        "-y",
        "@modelcontextprotocol/server-filesystem",
        "/path/to/allowed/folder"
      ]
    }
  }
}

Step 3: Restart Claude

Close and reopen Claude Desktop. The MCP server is now available.

Cursor Setup

Method 1: Settings UI

1. Open Cursor
2. Go to Settings → Features → MCP
3. Add your MCP server configuration

Method 2: Project Config

Create .cursor/mcp.json in your project:

{
  "servers": {
    "my-database": {
      "command": "npx",
      "args": ["-y", "@supabase/mcp-server"]
    }
  }
}

VS Code Setup

Add to your .vscode/settings.json:

{
  "mcp.servers": {
    "weather": {
      "command": "node",
      "args": ["./path/to/weather-server.js"]
    }
  }
}

Common MCP Server Configurations

Filesystem (Read/Write Files)

{
  "mcpServers": {
    "filesystem": {
      "command": "npx",
      "args": [
        "-y",
        "@modelcontextprotocol/server-filesystem",
        "/Users/yourname/projects"
      ]
    }
  }
}

Security: Only grant access to specific folders, not your entire computer.

Supabase (Database)

{
  "mcpServers": {
    "supabase": {
      "command": "npx",
      "args": ["-y", "@supabase/mcp-server"],
      "env": {
        "SUPABASE_URL": "your-project-url",
        "SUPABASE_KEY": "your-service-role-key"
      }
    }
  }
}

GitHub

{
  "mcpServers": {
    "github": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-github"],
      "env": {
        "GITHUB_TOKEN": "your-github-token"
      }
    }
  }
}

Testing Your Setup

After configuration:

1. Restart your application

2. Ask the AI to use the tool:

   • Filesystem: “List files in my projects folder”
   • Supabase: “Show me my database tables”
   • GitHub: “What are my recent commits?”

3. Check for errors in the application logs if it doesn’t work

Troubleshooting

”MCP server not found”

• Check the command path is correct
• Ensure Node.js is installed (node --version)
• Try running the command manually in terminal

”Permission denied”

• Check the allowed paths in your config
• Verify API keys/tokens are correct
• Some operations may need explicit approval

”Connection refused”

• Make sure no other process is using the same port
• Restart your AI application
• Check if the MCP server started successfully

Security Best Practices

1. Limit Access

Only grant access to specific folders/databases you need.

2. Use Environment Variables

Never hardcode API keys in config files:

{
  "env": {
    "API_KEY": "${MY_API_KEY}"  // Reads from system env
  }
}

3. Audit Regularly

Review what MCP servers are enabled occasionally.

4. Start Conservative

Enable only what you need. Add more later.

Next Steps

→ Popular MCP Servers — See what’s available
→ Automation Platforms — Combine MCPs with automation