MCP Excel Online
MCP server that integrates with OneDrive and Excel Workbooks via Microsoft Graph API for creating and modifying worksheets.
README
MCP Excel Online
An MCP (Model Context Protocol) server that integrates with your OneDrive and Excel Workbooks, to enable creating and modifying work sheets using the Microsoft Graph API.
Features
Resources
| Resource | Description |
|---|---|
drive://root/children |
List all items at the root of the configured OneDrive |
drive://drives/{drive_id}/root/{relative_path}/children |
List items in a folder by path |
Tools
| Tool | Description |
|---|---|
| List Sheets | List all worksheets in a workbook |
| Create Worksheet | Add a new worksheet to a workbook |
| Rename Worksheet | Rename an existing worksheet |
| Delete Worksheet | Delete a worksheet from a workbook |
| Copy Worksheet | Copy a worksheet from one workbook to another |
| Get Worksheet Data | Read cell values from a worksheet (full sheet or a range) |
| Get Worksheet Formulas | Read formulas from a worksheet (full sheet or a range) |
| Update Worksheet Data | Write values to a cell range in a worksheet |
Prerequisites
- Python 3.13+
uv(for local development)- Docker (for containerised deployment)
- A Microsoft Azure account with an Entra ID (Azure AD) app registration
Azure App Registration
Register your application by following the steps at Register your app with the Microsoft Identity Platform.
Permission Types
Application permissions are used when you don't need a user to log in to your app. The app performs tasks on its own and runs in the background using a client secret (service principal flow).
Required: APP_CLIENT_ID, APP_TENANT_ID, APP_CLIENT_SECRET
Delegated permissions (scopes) are used when your app requires a user to log in and interact with data related to that user in a session. Authentication is performed via the device code flow — the server will print a URL and a code to the console.
Required: APP_CLIENT_ID only
Environment Variables
| Variable | Required | Description |
|---|---|---|
APP_CLIENT_ID |
Always | The Application (client) ID of your Azure app registration |
APP_TENANT_ID |
Application permissions only | The Directory (tenant) ID of your Azure tenant |
APP_CLIENT_SECRET |
Application permissions only | A client secret generated for your app registration |
DRIVE_ID |
Required | OneDrive Drive ID to use as the default drive |
HOST |
Optional | Host address the server binds to (default: 0.0.0.0) |
PORT |
Optional | Port the server listens on (default: 8000) |
Create a .env file in the project root for local development:
APP_CLIENT_ID=your-client-id
APP_TENANT_ID=your-tenant-id # required for application permissions
APP_CLIENT_SECRET=your-client-secret # required for application permissions
DRIVE_ID=your-drive-id
Arguments
| Argument | Options | Default | Description |
|---|---|---|---|
--transport |
stdio, sse, streamable-http |
sse |
MCP transport protocol to use |
--graph-permission |
delegated, application |
delegated |
Microsoft Graph permission type (see above) |
Running Locally with uv
1. Clone the repository
git clone https://github.com/GokulSudheesh/mcp-excel-online.git
cd mcp-excel-online
2. Create your .env file with the required environment variables (see above).
3. Start the server
# SSE transport with delegated permissions (device code login)
uv run mcp-excel-online --transport sse --graph-permission delegated
# stdio transport with application permissions (service principal)
uv run mcp-excel-online --transport stdio --graph-permission application
# streamable-http transport
uv run mcp-excel-online --transport streamable-http
Running with Docker
1. Build the image
docker build -t mcp-excel-online .
2. Run the container
docker run --rm -p 8000:8000 \
-e APP_CLIENT_ID=your-client-id \
-e DRIVE_ID=your-drive-id \
mcp-excel-online
You can also pass a .env file:
docker run --rm -p 8000:8000 --env-file .env mcp-excel-online
The server is accessible at http://localhost:8000/sse (SSE transport).
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。