home / mcp / deploy mcp server

Deploy MCP Server

Tracks deployments across Vercel, Netlify, Cloudflare Pages, and more from your MCP client.

Installation

Add the following to your MCP client configuration file.

Configuration

{
  "mcpServers": {
    "alexpota-deploy-mcp": {
      "command": "npx",
      "args": [
        "-y",
        "deploy-mcp"
      ],
      "env": {
        "VERCEL_TOKEN": "YOUR_VERcel_TOKEN",
        "NETLIFY_TOKEN": "YOUR_NETLIFY_TOKEN",
        "CLOUDFLARE_TOKEN": "accountId:globalApiKey"
      }
    }
  }
}

deploy-mcp is a universal Deployment Tracker that runs alongside your AI assistant to monitor deployments across multiple platforms. It helps you see status, logs, and history from within your conversations, eliminating context-switching and dashboard hunting.

How to use

You interact with deploy-mcp through your MCP client. Start by running the server locally, then configure your client to connect to it so you can ask for deployment status, view logs, or watch progress in real time. You can track multiple platforms at once by providing tokens for each platform you want to monitor. Use the available commands to check status, fetch logs, monitor in real time, or list all projects across platforms.

How to install

Prerequisites: you need Node.js and npm installed on your machine.

Step 1: Install and run deploy-mcp locally using npx in a terminal.

npx deploy-mcp

Step 2: If you prefer a global installation for production use, install the package globally and start the server.

npm install -g deploy-mcp
deploy-mcp

Step 3: Configure your MCP client with platform tokens to enable multi-platform monitoring. The server reads tokens for each platform and enables the corresponding commands.

Platform configuration and usage

deploy-mcp supports multiple deployment platforms. You enable each by providing tokens in your configuration. The supported platforms shown are Vercel, Netlify, Cloudflare Pages, and GitHub Pages (coming soon). For each platform, you will get status, logs, history, and real-time monitoring once you supply the required token.

Example of enabling multiple platforms in one configuration: you supply tokens for each platform you want to track. The server then routes your queries to the appropriate platform APIs based on the project/site name in your command.

AI assistant configuration

You can connect deploy-mcp to several popular MCP clients. The configuration format generally mirrors a shared pattern: specify a command to start the MCP server, a set of arguments, and an env block with platform tokens.

Example integration snippet for a generic client would include the command to start the MCP server and environment variables for tokens.

Deployment status badges and webhooks

You can display live deployment status badges in your repositories using a shields.io-compatible endpoint. Real-time updates can be delivered via webhooks for Vercel and Netlify to push deployment events to your MCP server.

Note: Badges are public-repo-friendly; private repo support may be added in the future.

Security and best practices

Tokens and credentials are stored locally on your machine. There is no telemetry, and the codebase is open for auditing. Only minimum required permissions are requested from each platform.

Advanced configuration

Environment variables can also be set in your shell before starting deploy-mcp to avoid inline token usage.

Example: export VERCEL_TOKEN, NETLIFY_TOKEN, or CLOUDFLARE_TOKEN in your environment, then run the server.

Contributing and extending support

If you want to add support for a new platform, you should implement the required adapter, add the platform API client, and extend the configuration documentation so others can use it immediately.

Available tools

check_deployment_status

Get the latest deployment status or deployment history for a project across configured platforms.

watch_deployment

Monitor a deployment in real time while it progresses.

compare_deployments

Compare recent deployments to identify changes and trends.

get_deployment_logs

Fetch logs for a specific deployment or project.

list_projects

List all projects or sites available across configured platforms.