Skip to main content

​
Usage

superbox pull --name NAME --client CLIENT

​
Description

Pull an MCP server from the S3 registry and configure it for your AI client. Automatically writes the correct MCP configuration file for the specified client.

​
Options

​
--name
string
required
MCP server name to pull from registry
​
--client
string
required
Target AI client: vscode, cursor, windsurf, claude, or chatgpt

​
Supported Clients

VS Code

Configures .vscode/mcp.json

Cursor

Configures ~/.cursor/mcp.json

Windsurf

Configures Windsurf MCP settings

Claude

Configures Claude Desktop MCP

ChatGPT

Configures ChatGPT MCP settings

​
What It Does

1

Fetch from S3

Retrieves server metadata from S3 registry
2

Generate Config

Creates client-specific MCP configuration
3

Configure Endpoint

Points client to Lambda executor endpoint
4

Save Config

Writes configuration to client’s config file

​
Examples

superbox pull --name weather-mcp --client cursor

​
Example Output

$ superbox pull --name my-mcp --client cursor
Fetching server 'my-mcp' from S3 bucket...
Success!
Server 'my-mcp' added to Cursor MCP config
Location: ~/.cursor/mcp.json

​
Configuration Format

The generated configuration points to the Lambda executor: 
{
  "mcpServers": {
    "my-mcp": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-proxy"],
      "env": {
        "MCP_PROXY_URL": "https://lambda-url/my-mcp"
      }
    }
  }
}

## Next Steps

<CardGroup cols={2}>
  <Card title="Inspect Server" icon="eye" href="/cli/inspect">
    View server details
  </Card>
  <Card title="Run Locally" icon="play" href="/cli/run">
    Test server execution
  </Card>
</CardGroup>