Microsoft To Do MCP Server

Microsoft To Do MCP Server

Enables AI clients like Claude to read and manage Microsoft To Do task lists and tasks via the Microsoft Graph API, running locally over stdio with secure token storage.

Category
访问服务器

README

Microsoft To Do MCP Server

Microsoft-To-Do-MCP-Server

An MCP server that connects Claude (or any MCP-compatible client) to Microsoft To Do through the Microsoft Graph API — letting Claude read and manage your task lists and tasks on your behalf.

Runs entirely on your machine over stdio. No remote server ever sees your tasks or your tokens.

Features

Tool Description
todo_list_task_lists List all your task lists (Tasks, Groceries, etc.)
todo_create_task_list Create a new task list
todo_rename_task_list Rename an existing list
todo_delete_task_list Delete a list and all its tasks (destructive)
todo_list_tasks List tasks in a list, optionally filtered by status
todo_get_task Get the details of a single task
todo_create_task Create a task (title, notes, due date, importance)
todo_update_task Update fields on an existing task
todo_complete_task Mark a task as completed
todo_delete_task Delete a task (destructive)

Prerequisites

  • Node.js 18 or newer
  • A Microsoft account (personal or work/school)
  • Claude Desktop or another MCP-compatible client

1. Register an app in Azure AD

This server authenticates as your own Azure AD app registration, so you'll need to create one (free, takes a couple of minutes):

  1. Go to the Azure Portal → App registrations and click New registration.
  2. Give it a name (e.g. "Microsoft To Do MCP"), and under Supported account types choose an option that includes personal Microsoft accounts, unless you only intend to use a work/school account.
  3. Under Redirect URI, select platform "Mobile and desktop applications" and add:
    http://localhost:3000/callback
    
  4. Click Register, then copy the Application (client) ID from the overview page — you'll need it later.
  5. Go to Authentication and enable "Allow public client flows", then save.
  6. Go to API permissions → Add a permission → Microsoft Graph → Delegated permissions, and add:
    • Tasks.ReadWrite
    • offline_access
    • User.Read

2. Install & build

Clone this repository, then from the project folder:

git clone https://github.com/hoaaah/Microsoft-To-Do-MCP-Server.git
cd Microsoft-To-Do-MCP-Server
npm install
npm run build

This compiles the TypeScript source into dist/index.js.

3. Configure AI Client / Claude Desktop

Open your AI Client / Claude Desktop config file:

  • macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
  • Windows: %APPDATA%\Claude\claude_desktop_config.json

Add the following entry, replacing PATH_TO_THIS_FOLDER with the absolute path to where you cloned this repo, and MS_TODO_CLIENT_ID with the Application (client) ID from Azure:

{
  "mcpServers": {
    "microsoft-todo": {
      "command": "node",
      "args": ["PATH_TO_THIS_FOLDER/dist/index.js"],
      "env": {
        "MS_TODO_CLIENT_ID": "your-azure-client-id-here"
      }
    }
  }
}

Save the file and restart Claude Desktop.

4. First sign-in

The first time Claude calls one of the todo_* tools, your default browser will open automatically, asking you to sign in with your Microsoft account and approve the Tasks.ReadWrite permission. Once you approve, the token is cached locally at:

~/.microsoft-todo-mcp/token-cache.json

Subsequent sign-ins are silent — the cached refresh token is reused automatically as long as it hasn't expired or been revoked, so you won't need to log in again.

Revoking access

To force a fresh sign-in, delete the cache file:

~/.microsoft-todo-mcp/token-cache.json

To fully revoke the app's access to your account, go to:

Security notes

  • This server runs locally on your machine over stdio — there is no remote server storing or relaying your tokens.
  • todo_delete_task_list and todo_delete_task are destructive and cannot be undone. Claude will ask for confirmation before calling them.
  • Your Azure app's client ID is not a secret and is safe to keep in your local config, but you should not commit it to a public repository alongside real task data or tokens.

Troubleshooting

  • Browser doesn't open on first sign-in: make sure port 3000 is free, and that the redirect URI in your Azure app registration exactly matches http://localhost:3000/callback.
  • "Need admin approval" error: your organization may require an admin to consent to the requested permissions — contact your Microsoft 365 administrator, or use a personal Microsoft account instead.
  • Claude doesn't see the tools: double-check the path in claude_desktop_config.json points to the built dist/index.js, and that you restarted Claude Desktop after editing the config.

License

MIT

Creator

This MCP was created by and is maintained by Heru Arief Wijaya.

  • https://twitter.com/hoaaah
  • https://github.com/hoaaah

推荐服务器

Baidu Map

Baidu Map

百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。

官方
精选
JavaScript
Playwright MCP Server

Playwright MCP Server

一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。

官方
精选
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。

官方
精选
本地
TypeScript
VeyraX

VeyraX

一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。

官方
精选
本地
graphlit-mcp-server

graphlit-mcp-server

模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。

官方
精选
TypeScript
Kagi MCP Server

Kagi MCP Server

一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

Exa MCP Server

模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。

官方
精选