ntfy

You are setting up an MCP server that lets AI assistants send push notifications through ntfy.sh. This server provides a straightforward way to trigger real-time alerts from your AI workflows, delivering messages to devices via ntfy.sh channels and making it easy to stay informed about tasks and events in your applications.

How to use

You can use this MCP server from an MCP client to send notifications to any ntfy.sh channel. The server handles the integration with ntfy.sh, so you can focus on crafting timely alerts for important events, task completions, or automated workflow updates. Define the channel, message, and optional metadata like a title, priority, and tags to organize notifications.

How to install

Prerequisites you need on your machine before installing this MCP server:

• Node.js is required (preferably LTS version)
• npm (comes with Node.js)
• TypeScript tooling (ts-node, types for Node) to run the build and development workflow

Step-by-step installation and startup flow you should follow to get the MCP server up and running:

# Install dependencies
npm install --save-dev typescript @types/node ts-node

# Build the TypeScript code
npm run build

# Start the compiled MCP server
npm start

Configuration and usage notes

This server uses the public ntfy.sh service. You will send messages to ntfy.sh channels, and messages can include a title, a priority (1-5, with 5 as the highest), and tags to categorize notifications. Public channels do not require authentication, so anyone who knows a channel name can publish to it. Choose channel names carefully to avoid conflicts and unintended access.

Additional content

Configuration basics you should know:

• The server is intended to be run locally or in your environment and exposes a simple interface for sending messages to ntfy.sh channels.

Available tools

send_message

Sends a message to a specified ntfy.sh channel with optional title, priority, and tags. Returns a success/failure result and publishes the notification to the chosen channel.