Outlook MCP Python
A Python-based MCP server for Microsoft Outlook integration using Microsoft Graph API, enabling email reading/sending, calendar management, and contact operations through Claude Desktop.
README
Outlook MCP Python
A Python-based Microsoft Outlook integration using Microsoft Graph API, built with FastAPI and MCP framework.
Overview
This project provides a server implementation for Microsoft Outlook integration using the Microsoft Graph API. It includes authentication handling, mail operations, and various utility functions for working with Outlook data.
Features
- Microsoft Graph API integration
- OAuth2 authentication flow
- Mail operations (send, read, manage)
- FastAPI-based server implementation
- Environment-based configuration
- Logging functionality
Project Structure
auth/- Authentication related modulesmail/- Mail operation modulesutils/- Utility functionsmain.py- Main application entry pointserver.py- Server configurationconfig.py- Configuration settingslogger.py- Logging configurationoutlook_auth_server.py- Outlook authentication server implementation
Prerequisites
- Python 3.8 or higher
- Microsoft Azure account with registered application
- Microsoft Graph API access
Installation
- Clone the repository:
git clone [repository-url]
cd mcp-outlook-python
- Create and activate a virtual environment:
python -m venv venv
source venv/bin/activate # On Windows: venv\Scripts\activate
- Install dependencies:
pip install -r requirements.txt
Azure App Registration & Configuration
To use this MCP server you need to first register and configure an app in Azure Portal. The following steps will take you through the process of registering a new app, configuring its permissions, and generating a client secret.
App Registration
- Open Azure Portal in your browser
- Sign in with a Microsoft Work or Personal account
- Search for or cilck on "App registrations"
- Click on "New registration"
- Enter a name for the app, for example "Outlook MCP Server"
- Select the "Accounts in any organizational directory and personal Microsoft accounts" option
- In the "Redirect URI" section, select "Web" from the dropdown and enter "http://localhost:3333/auth/callback" in the textbox
- Click on "Register"
- From the Overview section of the app settings page, copy the "Application (client) ID" and enter it as the MS_CLIENT_ID in the .env file as well as the OUTLOOK_CLIENT_ID in the claude-config-sample.json file
App Permissions
- From the app settings page in Azure Portal select the "API permissions" option under the Manage section
- Click on "Add a permission"
- Click on "Microsoft Graph"
- Select "Delegated permissions"
- Search for the following permissions and slect the checkbox next to each one
- offline_access
- User.Read
- Mail.Read
- Mail.Send
- Calendars.Read
- Calendars.ReadWrite
- Contacts.Read
- Click on "Add permissions"
Client Secret
- From the app settings page in Azure Portal select the "Certificates & secrets" option under the Manage section
- Switch to the "Client secrets" tab
- Click on "New client secret"
- Enter a description, for example "Client Secret"
- Select the longest possible expiration time
- Click on "Add"
- Copy the secret value and enter it as the MS_CLIENT_SECRET in the .env file as well as the OUTLOOK_CLIENT_SECRET in the claude-config-sample.json file
Configuration
- Create a
.envfile in the root directory with the following variables:
MS_CLIENT_ID=your-ms-client-id
MS_CLIENT_SECRET=your-ms-client-secret
MS_AUTH_SERVER_URL=your-ms-auth-server-url
- Update the configuration in
config.pyas needed.
Usage with Claude Desktop
- Copy the sample configuration from
claude-config-sample.jsonto your Claude Desktop configuration - Restart Claude Desktop
- Authenticate with Microsoft using the
authenticatetool - Use the email tools to manage your Outlook account
Authentication Flow
- Start a local authentication server on port 3333 (using
outlook-auth-server.js) - Use the
authenticatetool to get an authentication URL - Complete the authentication in your browser
- Tokens are stored in
~/.outlook-mcp-tokens.json
Troubleshooting
- Authentication Issues: Check the token file and authentication server logs
- API Call Failures: Check for detailed error messages in the response
Dependencies
- fastapi
- uvicorn
- python-dotenv
- pydantic
- pydantic-settings
- requests
- aiohttp
- mcp
推荐服务器
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。
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。
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。
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。
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。
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。
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。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
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。