Install the MCP server

Connect your AI assistant to Andromesh Labs and try your first MCP tool. Our hosted demo returns a personal greeting. No download, account, or API key is needed.

Your server connection

Server URL

https://www.andromeshlabs.com/api/mcp

Transport: Streamable HTTP
Authentication: None
Available tool: hello_world

1. Add the server to your assistant

Use an MCP client that supports remote servers over Streamable HTTP. Open its MCP or integrations settings, add a remote server named Andromesh Labs, and paste the server URL above. Save the connection and enable its tools if prompted.

The exact settings depend on your client. This URL is an MCP endpoint; opening it in a browser does not run the greeting tool.

Using Cursor

Create .cursor/mcp.json in your project, or use ~/.cursor/mcp.json for all projects. Add this configuration. If you already have servers configured, add the andromesh-labs entry to your existing mcpServers object.

{
  "mcpServers": {
    "andromesh-labs": {
      "url": "https://www.andromeshlabs.com/api/mcp"
    }
  }
}

See the Cursor MCP documentation for current configuration options.

2. Try a greeting

Start a conversation with your assistant and ask:

Use the Andromesh Labs hello_world tool with the name Ada.

Approve the tool call if your assistant asks. The tool accepts a required, non-empty name string and returns:

Hello, Ada

Replace Ada with your name to make it your own.

If the connection does not work

Check that the URL includes https://www.andromeshlabs.com/api/mcp.
Choose Streamable HTTP if your client asks for a transport.
Refresh the server connection and confirm hello_world appears in the tool list.
Check that your client allows the tool and that the name is a non-empty string.

From a greeting to a useful workflow

This server is a small connection demo with one greeting tool. Learn how MCP works, or explore our MCP development services to connect an assistant to your business.

Talk to a human