Step-by-Step Guide: Creating an MCP Server with Vectorize for Invoice Data Analysis

Step 3: Connect Claude Desktop to the MCP Server

  1. Open the Claude Desktop App, click the hamburger menu, select File, then go to Settings.

  1. Select the Developer menu and click "Edit Config".

  1. Select the file name: claude_desktop_config.json and double click on the file to open it.

  1. Copy the configuration file and past it into the file: claude_desktop_config.json and save it. 

{
  "mcpServers": {
    "vectorize-search": {
      "command": "C:\\mcprag\\vectorize-search\\.venv\\Scripts\\python.exe",
      "args": [
        "C:\\mcprag\\vectorize-search\\server.py"
      ],
      "host": "127.0.0.1",
      "port": 8080,
      "timeout": 30000
    }
  }
}

  1. Execute the MCP server by running the following command in your terminal: python server.py If successful, you'll see this message: Starting MCP server 'vectorize-search' on 127.0.0.1:8080 ...

Before you execute the MCP server, please make sure that activate your virtual environment with this command:

PS C:\mcptool\vectorize-search>.venv\Scripts\activate (For Windows) or source .venv\bin\activate (For Mac) and press enter

  1. Open the Claude Desktop App, click the hamburger menu, select "File," then go to "Settings." Select "Developer" and choose vectorize-search. If successful, you will see a "running" status.

  1. Exit the Developer menu. You will then see a hammer icon with "1 MCP tool available" in the chat box.

  1. Click on the hammer icon to see the vectorize_search tool available.