mcpField
F
FastMCP Todo List Server

FastMCP Todo List Server

Local
github.com/DustyPolkProductivity1Updated 77w ago

About FastMCP Todo List Server

FastMCP Todo List Server is a todo management server for people organizing tasks through an MCP client such as Claude Desktop. It combines filtering, search, backups, and statistics with JSON-file storage.

Key features

  1. 01Create, update, and delete todos
  2. 02Filter, sort, and paginate tasks
  3. 03Search titles and descriptions
  4. 04Create automatic and manual backups
  5. 05View todo statistics
  6. 06Configure data and backup paths

Use cases

  1. 01Tracking tasks by status, priority, tags, or due date
  2. 02Finding a task by its title or description
  3. 03Backing up todo data before changes
  4. 04Reviewing task statistics

MCP support

Exposes todo management tools to MCP clients, including task CRUD, filtering, search, backup creation, and statistics.

Configuration

Local server config is the same for every client. Copy it into your client’s MCP config file.

No install method listed yet

We haven’t collected a package or remote endpoint for this server. Follow the instructions in the project README.

Repository

Tools

No tool list yet

Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.

Details

From the project README

FastMCP Todo List Server

A powerful, feature-rich todo list server built with FastMCP that provides a comprehensive set of tools for managing todos. This server can be used with Claude Desktop to create, manage, and organize your tasks efficiently.

Features

  • Rich Todo Management: Create, read, update, and delete todos with extensive metadata
  • Advanced Filtering & Sorting: Filter by status, priority, tags, due dates, and more
  • Search Capability: Search todos by title and description
  • Automatic Backups: Automatic backups before any data changes
  • Manual Backups: Create manual backups on demand
  • Statistics: Get detailed statistics about your todos
  • Data Validation: Comprehensive validation for all inputs
  • Caching: Efficient caching to reduce file I/O
  • Configurable: Configurable file paths via environment variables

Installation

  1. Clone this repository:

    git clone https://github.com/DustyPolk/todo_mcp.git
    cd todo_mcp
    
  2. Create and activate a virtual environment:

    python -m venv .venv
    # On Windows
    .venv\Scripts\activate
    # On macOS/Linux
    source .venv/bin/activate
    
  3. Install the required packages:

    pip install fastmcp aiofiles
    

Running the Server

Run the server with:

python todolist_server.py

The server will start and be available for Claude Desktop to connect to.

Environment Variables

You can configure the server using the following environment variables:

  • TODO_FILE: Path to the JSON file where todos are stored (default: todos.json in the same directory as the script)
  • TODO_BACKUP_DIR: Path to the directory where backups are stored (default: backups in the same directory as the script)

Available Tools

The server provides the following tools:

add_todo

Add a new todo item.

Parameters:

  • title: The title of the todo (required, max 100 chars)
  • description: A detailed description (optional, max 1000 chars)
  • due_date: Due date in YYYY-MM-DD format (optional)
  • status: Status of the todo (optional, one of: pending, in_progress, done, cancelled)
  • priority: Priority level (optional, one of: low, medium, high, critical)
  • tags: List of tags for categorization (optional)
list_todos

List todos with filtering, sorting and pagination.

Parameters:

  • status: Filter by status (optional, one of: pending, in_progress, done, cancelled)
  • priority: Filter by priority (optional, one of: low, medium, high, critical)
  • search: Search in title and description (optional)
  • tag: Filter by tag (optional)
  • due_date_filter: Filter by due date (optional, one of: overdue, today, upcoming, no_date)
  • sort_by: Field to sort by (optional, default: id)
  • sort_order: Sort order (optional, one of: asc, desc, default: asc)
  • limit: Maximum number of todos to return (optional, default: 100)
  • offset: Number of todos to skip (optional, default: 0)
get_todo

Get a todo by id.

Parameters:

  • todo_id: The ID of the todo to retrieve
update_todo

Update a todo by id.

Parameters:

  • todo_id: The ID of the todo to update
  • title: New title (optional)
  • description: New description (optional)
  • due_date: New due date in YYYY-MM-DD format (optional)
  • status: New status (optional, one of: pending, in_progress, done, cancelled)
  • priority: New priority (optional, one of: low, medium, high, critical)
  • tags: New list of tags (optional)
complete_todo

Mark a todo as complete by id.

Parameters:

  • todo_id: The ID of the todo to mark as complete
delete_todo

Delete a todo by id.

Parameters:

  • todo_id: The ID of the todo to delete
batch_delete_todos

Delete multiple todos by ids.

Parameters:

  • todo_ids: List of todo IDs to delete
create_backup

Create a manual backup of the todos file.

Parameters: None

get_statistics

Get statistics about todos.

Parameters: None

Data Structure

Todos are stored with the following structure:

{
  "id": 1,
  "title": "Complete project report",
  "description": "Finish the quarterly project report",
  "due_date": "2025-05-20",
  "status": "pending",
  "priority": "high",
  "tags": ["work", "quarterly"],
  "created_at": "2025-05-09T12:00:00.000000",
  "updated_at": "2025-05-09T12:00:00.000000"
}

License

MIT

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.