Ntfy MCP Server for LLM Notifications

An MCP (Model Context Protocol) server designed to interact with the ntfy push notification service. It enables LLMs and AI agents to send notifications to your devices with extensive customization options.

Table of Contents

• Overview
• Features
• Quick Start
• Installation
• Configuration
• Project Structure
• Tools
• Resources
• Use Cases
• Available Scripts
• Contributing
• License

Overview

This server implements the Model Context Protocol (MCP), enabling standardized communication between LLMs and external systems. Specifically, it provides an interface to the ntfy push notification service.

┌───────────┐      ┌───────────┐      ┌───────────┐      ┌─────────┐
│ LLM Agent │ ────▶│ Ntfy MCP  │ ────▶│ Ntfy      │ ────▶│ Your    │
│ (Claude)  │      │ Server    │      │ Service   │      │ Devices │
└───────────┘      └───────────┘      └───────────┘      └─────────┘

Features

• MCP Server Implementation: Built using the @modelcontextprotocol/sdk for seamless integration with LLM agents.
• Ntfy Integration: Provides a tool (send_ntfy) to send notifications with support for:
• Message prioritization (1-5 levels)
• Emoji tags
• Clickable actions and buttons
• File attachments
• Delayed delivery
• Markdown formatting
• Resource Exposure: Exposes the configured default ntfy topic as an MCP resource.
• TypeScript: Modern, type-safe codebase with comprehensive type definitions.
• Structured Logging: Uses winston and winston-daily-rotate-file for detailed and rotatable logs.
• Configuration Management: Uses dotenv for easy environment-based configuration.
• Utility Scripts: Includes scripts for cleaning build artifacts and generating directory structure documentation.
• Error Handling & Security: Implements robust error handling, input sanitization (sanitize-html), and security filters (xss-filters).

Quick Start

1. Prerequisites:
2. Node.js (v16+)
3. npm or yarn
4. An MCP-compatible client (Claude Desktop, Cline, etc.)
5. Install and Run:
   ```shell
   # Option 1: Install via npm
   npm install -g ntfy-mcp-server

# Option 2: Clone repository and build
git clone https://github.com/cyanheads/ntfy-mcp-server.git
cd ntfy-mcp-server
npm install
npm run build

# Create .env file (optional but recommended)
cp .env.example .env
# Edit .env to set NTFY_DEFAULT_TOPIC

# Start the server
npm start
`` 3. **Add to MCP Client Settings:** Add the server to your MCP client settings file (see [Configuration](#configuration)). 4. **Use the tool:** Once connected, you can use thesend_ntfy` tool to send notifications.

Installation

Option 1: NPM Package (Recommended)

1. Install the package globally:
   shell npm install -g ntfy-mcp-server
2. Or install locally in your project:
   shell npm install ntfy-mcp-server

Option 2: From Source

1. Clone the repository:
   shell git clone https://github.com/cyanheads/ntfy-mcp-server.git cd ntfy-mcp-server
2. Install dependencies:
   shell npm install
3. Build the project:
   shell npm run build

Configuration

Environment Variables

Create a .env file in the project root based on .env.example:

# Ntfy Configuration
NTFY_BASE_URL=https://ntfy.sh  # Optional: Base URL of your ntfy instance
NTFY_DEFAULT_TOPIC=your_default_topic # Optional: Default topic if none specified in requests

# Application Configuration
LOG_LEVEL=info # Optional: Logging level (debug, info, warn, error)
NODE_ENV=development # Optional: Environment (development, production)

MCP Client Settings

For Cline VSCode Extension

Add the following configuration to your Cline MCP settings file:

{
  "mcpServers": {
    "ntfy": {
      "command": "ntfy-mcp-server",
      "env": {
        "NTFY_BASE_URL": "https://ntfy.sh",
        "NTFY_DEFAULT_TOPIC": "your_default_topic",
        "LOG_LEVEL": "info",
        "NODE_ENV": "production"
      }
    }
  }
}

For Claude Desktop App

Add the following configuration to your Claude Desktop config file:

{
  "mcpServers": {
    "ntfy": {
      "command": "ntfy-mcp-server",
      "env": {
        "NTFY_BASE_URL": "https://ntfy.sh",
        "NTFY_DEFAULT_TOPIC": "your_default_topic",
        "LOG_LEVEL": "info",
        "NODE_ENV": "production"
      }
    }
  }
}

Ntfy Setup

1. Install the ntfy app on your devices from ntfy.sh or the app stores.
2. Subscribe to your topic in the app.
3. Use the same topic in your MCP server configuration.

Project Structure

.
├── .env.example            # Example environment variables
├── .gitignore              # Git ignore patterns
├── LICENSE                 # Project license (Apache-2.0)
├── package.json            # Project metadata and dependencies
├── tsconfig.json           # TypeScript compiler configuration
├── docs/
│   └── tree.md             # Auto-generated directory structure
├── logs/                   # Runtime logs (created automatically)
├── scripts/                # Utility scripts
│   ├── clean.ts            # Cleans build artifacts and logs
│   └── tree.ts             # Generates the docs/tree.md file
└── src/                    # Source code
    ├── index.ts            # Main server entry point
    ├── config/             # Configuration loading
    ├── mcp-server/         # MCP server logic, tools, and resources
    │   ├── resources/      # MCP resource implementations
    │   ├── tools/          # MCP tool implementations
    │   └── utils/          # MCP-specific utilities
    ├── services/           # External service integrations (ntfy)
    ├── types-global/       # Global type definitions
    └── utils/              # General utility functions

Tools

send_ntfy

Sends a notification message via the ntfy service.

Key Arguments:

Example Usage:

// Basic notification
{
  "topic": "alerts",
  "message": "The task has completed successfully."
}

// Advanced notification
{
  "topic": "alerts",
  "title": "System Alert",
  "message": "CPU usage has exceeded 90% for 5 minutes.",
  "tags": ["warning", "computer"],
  "priority": 4,
  "click": "https://server-dashboard.example.com",
  "actions": [
    {
      "id": "view",
      "label": "View Details",
      "action": "view",
      "url": "https://server-dashboard.example.com/cpu"
    },
    {
      "id": "restart",
      "label": "Restart Service",
      "action": "http",
      "url": "https://api.example.com/restart-service",
      "method": "POST",
      "headers": {
        "Authorization": "Bearer token123"
      }
    }
  ],
  "markdown": true
}

Example Response:

{
  "success": true,
  "id": "5ZFY362156Sa",
  "topic": "ATLAS",
  "time": 1743064235,
  "expires": 1743496235,
  "message": "This is a test message from the README verification process",
  "title": "README Testing"
}

Resources

Direct Resources

ntfy://default

• Description: Returns the default ntfy topic configured in the server's environment variables (NTFY_DEFAULT_TOPIC).
• Usage: Useful for clients to discover the primary topic without needing prior configuration.
• Example Response:
  json { "defaultTopic": "ATLAS", "timestamp": "2025-03-27T08:30:25.619Z", "requestUri": "ntfy://default", "requestId": "0da963d0-30e0-4dbc-bb77-4bf2dee14484" }

Resource Templates

ntfy://{topic}

• Description: Returns information about a specific ntfy topic.
• Parameters: topic - The name of the ntfy topic.
• Example Response:
  json { "topic": "ATLAS", "timestamp": "2025-03-27T08:30:30.038Z", "requestUri": "ntfy://ATLAS", "requestId": "31baf1df-278f-4fdb-860d-019f156a72b0" }

Use Cases

1. Long-running Task Notifications - Get notified when tasks like database backups, code generation, or data processing complete.
2. Scheduled Reminders - Set delayed notifications for future events or reminders.
3. Alert Systems - Set up critical alerts for monitoring systems or important events.
4. Mobile Notifications from LLMs - Allow LLMs to send notifications directly to your phone.
5. Multi-step Process Updates - Receive updates as different stages of a complex process complete.

Usage Examples

Basic Notification

<use_mcp_tool>
<server_name>ntfy-mcp-server</server_name>
<tool_name>send_ntfy</tool_name>
<arguments>
{
  "topic": "updates",
  "title": "Task Completed",
  "message": "Your requested data analysis has finished",
  "tags": ["check"]
}
</arguments>
</use_mcp_tool>

Rich Notification with Actions

<use_mcp_tool>
<server_name>ntfy-mcp-server</server_name>
<tool_name>send_ntfy</tool_name>
<arguments>
{
  "topic": "alerts",
  "title": "Critical Error Detected",
  "message": "The application has encountered a critical error.\n\n**Error Code**: E123\n\n**Details**: Database connection failed",
  "tags": ["warning", "skull"],
  "priority": 5,
  "actions": [
    {
      "id": "view",
      "label": "View Logs",
      "action": "view",
      "url": "https://logs.example.com"
    },
    {
      "id": "restart",
      "label": "Restart Service",
      "action": "http",
      "url": "https://api.example.com/restart",
      "method": "POST"
    }
  ],
  "markdown": true
}
</arguments>
</use_mcp_tool>

Available Scripts

• npm run build: Compiles the TypeScript source code to JavaScript in the dist/ directory.
• npm run clean: Removes the dist/ directory and cleans the contents of the logs/ directory.
• npm run rebuild: Runs clean and then build.
• npm run tree: Generates a directory tree representation in docs/tree.md.
• npm start: Runs the compiled server from the dist/ directory using Node.js.
• npm run watch: Tails the combined log file (logs/combined.log) for real-time monitoring.

Contributing

Contributions are welcome! Please feel free to submit pull requests or open issues to improve the project.

1. Fork the repository.
2. Create a feature branch (git checkout -b feature/your-feature).
3. Commit your changes (git commit -m 'Add some feature').
4. Push to the branch (git push origin feature/your-feature).
5. Create a new Pull Request.

Development Best Practices

• Follow TypeScript best practices and maintain strong typing
• Write tests for new functionality
• Keep dependencies up to date
• Follow the existing code style and patterns

License

This project is licensed under the Apache-2.0 License. See the LICENSE file for details.

Acknowledgements

• ntfy.sh for providing the notification service
• Model Context Protocol for enabling LLM-to-tool connections
• All contributors and users of this project