Apple Reminders MCP Server

A Model Context Protocol (MCP) server that enables AI assistants to interact with Apple Reminders on macOS. This server provides tools to create, read, list, and delete reminders programmatically through AppleScript.

Features

• ✅ Create reminders with due dates, times, notes, and locations

• 📋 List all reminder lists

• 🔍 Fetch reminders from specific lists (completed or pending)

• 🗑️ Delete reminders by name

• 🤖 Compatible with Claude Desktop and other MCP clients

• 🍎 Native macOS integration via AppleScript

• 🔒 Input sanitization to prevent AppleScript injection

Related MCP server: Things MCP Server

Demo

https://github.com/user-attachments/assets/d36e0df3-ff4b-4f48-bb29-41ddb5483c6b

Prerequisites

• macOS (required for AppleScript)

• Python 3.10 or higher

• Apple Reminders app

Installation

Option 1: Install from PyPI (Recommended)

pip install apple-reminders-mcp

Option 2: Install from source

git clone https://github.com/shreyanshjain05/apple_reminder_mcp_server.git
cd apple_reminder_mcp_server
pip install -e .

Usage

Running the Server

If installed via pip:

apple-reminders-mcp

Or run directly:

python -m apple_reminders_mcp.server

Testing with MCP Inspector

npx @modelcontextprotocol/inspector python3 server.py

This will:

• Start a proxy server on localhost:6277

• Open the inspector interface in your browser at http://localhost:6274

• Display a session token for authentication

Integration with Claude Desktop

Add to your Claude Desktop configuration file:

macOS: ~/Library/Application Support/Claude/claude_desktop_config.json

{
  "mcpServers": {
    "apple-reminders": {
      "command": "python3",
      "args": ["/path/to/apple_reminder_mcp_server/server.py"]
    }
  }
}

Note: Replace /path/to/ with the actual path where you cloned the repo. If you get "Server disconnected" errors, use the full path to your Python executable (run which python3 to find it).

Available Tools

1. create_reminder

Creates a new reminder in Apple Reminders.

Parameters:

• title (required): The title of the reminder

• due_date (required): Due date (e.g., "2024-12-25", "tomorrow", "next Friday")

• due_time (optional): Time in HHMMSS format (default: 090000 for 9 AM)

• notes (optional): Additional notes/body text

• list_name (optional): Target list name (default: "Reminder Created using Agent")

• location (optional): Location for location-based reminders

Example:

{
  "title": "Team Meeting",
  "due_date": "tomorrow",
  "due_time": "140000",
  "notes": "Discuss Q1 goals",
  "list_name": "Work"
}

2. get_reminder

Fetches reminders from a specific list.

Parameters:

• list_name (required): Name of the reminder list

• completed (optional): Whether to fetch completed reminders (default: false)

• limit (optional): Maximum number of reminders to return (default: 20)

3. list_reminder_lists

Returns all available reminder lists. No parameters required.

4. delete_reminder

Deletes a reminder by name.

Parameters:

• name (required): The exact name of the reminder to delete

• list_name (optional): Specific list to search in (searches all lists if not provided)

Troubleshooting

Common Issues

1. Permission errors with AppleScript

   • Grant Terminal/IDE permission to control Reminders in System Preferences → Security & Privacy → Automation

2. "No reminder list found" error

   • Make sure the list name exactly matches (case-sensitive)

   • Use list_reminder_lists to see available lists

3. Date parsing errors

   • Use standard formats like "YYYY-MM-DD" or natural language like "tomorrow"

Development

Running Tests

pip install -e ".[dev]"
pytest tests/ -v

Project Structure

apple_reminder_mcp_server/
├── .github/workflows/
│   ├── test.yml           # CI testing
│   └── publish.yml        # PyPI publishing
│── __init__.py
├── pyproject.toml         # Package configuration
├── server.py              # Main MCP server implementation
└── README.md 
├── tests/
│   ├── __init__.py
│   └── test_server.py     # Unit tests

Contributing

See CONTRIBUTING.md for guidelines.

License

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

Acknowledgments

• Built with FastMCP

• Uses AppleScript for native macOS integration

• Compatible with the Model Context Protocol