OneNoteMCP
Enables interaction with Microsoft OneNote via the Microsoft Graph API, allowing users to list notebooks and retrieve page content. It supports both personal and organization notebooks with credential caching for efficient authentication.
README
Forked from ZubeidHendricks/azure-onenote-mcp-server see that page for the original README.md
About
An MCP for OneNote. Forked and revised extensively, now works with personal notebooks. Updated to the latest MCP API. Supports caching credentials.
Improvements
Functionality
- Fixed
getPageContentfunctionality by implementing ReadableStream - Successfully implemented API calls against personal notebooks
- Updated index and tools to use the latest MCP version
Code Architecture
- Simplified codebase by removing classes that weren't necessary
- Many unnecessary files removed
Authentication & Performance
- Implemented disk caching for authentication credentials
- Updated package dependencies
- Added dotenv support for direct testing capabilities
Azure App Registration
- Go to Azure Portal and navigate to App registrations
- Create a new registration for OneNoteMCP
- Add Microsoft Graph API permissions:
- Notes.Read
- Notes.Read.All
- (optionally add Notes.Write permissions, but there is a risk of losing your notes)
- Create a client secret
- Copy the client ID/secret for configuration
- In the manifest set
signInAudiencetoAzureADandPersonalMicrosoftAccount
MCP Server setup
Put the client ID/secret n your .env file locally as well as in Claude desktop.
The AUTH_CACHE_DIR should be the directory where you want the generated API access token to be stored.
AZURE_CLIENT_ID=\
AZURE_CLIENT_SECRET=
AUTH_CACHE_DIR="C:\\git\\azure-onenote-mcp-server\\.cache"
Run server manually first to generate the credentials that will be cached and used by Claude desktop when it runs the server later on.
npm install
npm build
npm run
You can test the API works by running this command directly once the server starts up and logs in.
{"jsonrpc": "2.0", "id": 3, "method": "tools/call", "params": {"name": "listNotebooks", "arguments": {}}}
Claude MCP config
{
"mcpServers": {
"onenote": {
"command": "node",
"args": ["C:\\git\\azure-onenote-mcp-server\\dist\\index.js"],
"env": {
"AZURE_CLIENT_ID": "...",
"AZURE_CLIENT_SECRET": "...",
"AUTH_CACHE_DIR": "C:\\git\\azure-onenote-mcp-server\\.cache"
}
}
}
}
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。