home / mcp / ops mcp server

Ops MCP Server

Real MCP WebSocket server for Gmail operations management, enabling ops reports and task management via Agent Builder.

Installation

Add the following to your MCP client configuration file.

Configuration

{
  "mcpServers": {
    "johnohhh1-ops-mcp-server": {
      "url": "wss://ops-mcp-server.up.railway.app",
      "headers": {
        "PORT": "3334",
        "AI_SECRET_TOKEN": "YOUR_TOKEN"
      }
    }
  }
}

Ops MCP Server provides a WebSocket-based MCP server for Gmail operations management. It connects to Agent Builder to fetch ops reports and manage tasks, enabling you to automate workflows and monitor tasks from a single endpoint.

How to use

You will connect an MCP client to the Ops MCP Server using a WebSocket URL. Start the server locally to run on ws://localhost:3334, or deploy to Railway to obtain a public WebSocket URL. From an MCP client, use the provided endpoints to fetch ops reports and mark tasks as done. You can configure your client to point to either the local URL or the remote Railway URL as your MCP endpoint. When you connect, you will gain access to tools that fetch operational reports and allow you to complete tasks directly from the client.

Key actions you can perform include fetching the latest ops reports and marking tasks as completed by ID. Ensure you supply the required secret token for authentication if your deployment environment requires it. If you are running locally, the server uses the WebSocket port 3334 by default, and you can override this with an environment variable if needed.

How to install

Prerequisites you need before installing: Node.js and npm. Ensure Node and npm are available on your machine.

# Install dependencies
npm install

# Copy and configure environment
cp .env.example .env
# Edit .env and add your AI_SECRET_TOKEN

Run the server locally to start in development mode. The server will listen on the default WebSocket endpoint. If you want to expose it locally, start the server with the command below.

npm start
# Server runs on ws://localhost:3334

Additional notes

Deployment options include publishing to Railway. You will obtain a public WebSocket URL after deployment, and you can configure an environment variable named AI_SECRET_TOKEN in the deployment environment.

Environment variables you should be aware of include AI_SECRET_TOKEN for authentication and PORT to configure the WebSocket port if you need a custom value. The default port is 3334.

Available tools

fetch_ops_report

Fetches Ops reports including HotSchedules 911 items and Brinker deadlines from the connected Gmail operations context.

mark_task_done

Marks a specified task as complete by its ID within the MCP workflow.