ms-graph-dev

ms-graph-dev

Helps construct and validate Microsoft Graph REST API calls with on-demand tool loading, providing request details, required permissions, and code examples for categories like users, files, mail, calendar, groups, notes, tasks, sites, and subscriptions.

Category
访问服务器

README

Microsoft Graph Dev MCP

An MCP server that helps you construct and validate Microsoft Graph REST API calls — no authentication required in the server itself.

Tools are loaded on demand by resource category. Ask about SharePoint and the sites tools appear. Ask about files and the OneDrive tools appear. The server starts lean and grows with your needs, and always knows the required permissions for every operation.

What it does

  • Constructs valid Graph API request URLs, methods, headers, and bodies
  • Validates required and optional parameters
  • Returns required Microsoft Graph permissions (delegated and application) for every operation
  • Returns ready-to-use fetch code examples
  • Links to official Microsoft documentation for every operation
  • Explains cross-cutting concepts: pagination, OData queries, throttling, delta sync, batching, and auth
  • Loads resource categories on demand — only what you need

Tools

Always available

These tools are loaded immediately — no setup required.

Tool Description
list_categories List all resource categories and which are currently loaded
load_category Load tools for a category; triggers tools/list_changed
search_graph_api Search Graph API endpoints by keyword; returns a suggestedCategory to load
graph_build_batch Build a valid /$batch request body from up to 20 operations
graph_explain_pagination How @odata.nextLink works; iterate all pages of results
graph_explain_odata $filter, $select, $expand, $orderby, $count, $search with examples
graph_explain_throttling 429 handling, Retry-After, exponential backoff pattern
graph_explain_delta Delta tokens, change tracking, initial sync vs incremental sync
graph_explain_batch JSON batching, dependsOn, response handling
graph_explain_permissions Delegated vs application permissions, consent flows, token acquisition

Resource categories (loaded on demand)

Category Tools What you get
users 8 Get, list, create, update, delete users; manager; direct reports; delta sync
files 11 OneDrive/SharePoint document library: list, get, upload (<4MB and resumable), create folder, delete, move, copy, search, download URL, delta sync
mail 9 List and get messages, send, create draft, reply, delete, move, list folders, delta sync
calendar 8 List, get, create, update, delete events; find meeting times; get free/busy schedule; delta sync
groups 8 List, get, create, delete groups; list and manage members and owners
notes 8 OneNote notebooks, sections, pages; create and read content
tasks 10 Planner plans and tasks (CRUD); Microsoft To Do lists and tasks
sites 12 SharePoint sites; lists; list items (CRUD); columns
subscriptions 5 Create, list, get, delete, and renew webhook change notification subscriptions

Example tool output

Every tool returns a structured object with the full request details and required permissions:

{
  "endpoint": "https://graph.microsoft.com/v1.0/sites/{siteId}/lists/{listId}/items?expand=fields",
  "method": "GET",
  "headers": {
    "Authorization": "Bearer {token}"
  },
  "pathParams": { "siteId": "contoso.sharepoint.com,abc123", "listId": "list456" },
  "queryParams": { "expand": "fields" },
  "body": null,
  "description": "List items in list list456.",
  "docsUrl": "https://learn.microsoft.com/en-us/graph/api/listitem-list",
  "codeExample": "const response = await fetch('...', { method: 'GET', headers: { Authorization: 'Bearer {token}' } });\nconst data = await response.json();",
  "requiredPermissions": {
    "delegated": ["Sites.Read.All"],
    "application": ["Sites.Read.All"]
  },
  "notes": "Column values are returned under the 'fields' property. This request already includes ?expand=fields. Without it the items array would contain only metadata, not column data."
}

Installation

The recommended way to run this server is via npx — no local install needed.

npx -y @stubbedev/ms-graph-dev-mcp

Claude Desktop

Edit ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows):

{
  "mcpServers": {
    "ms-graph-dev": {
      "command": "npx",
      "args": ["-y", "@stubbedev/ms-graph-dev-mcp"]
    }
  }
}

Claude Code (CLI)

claude mcp add ms-graph-dev -- npx -y @stubbedev/ms-graph-dev-mcp

Or add to your project's .mcp.json:

{
  "mcpServers": {
    "ms-graph-dev": {
      "command": "npx",
      "args": ["-y", "@stubbedev/ms-graph-dev-mcp"]
    }
  }
}

Cursor

Open Settings → MCP and add a new server, or edit ~/.cursor/mcp.json:

{
  "mcpServers": {
    "ms-graph-dev": {
      "command": "npx",
      "args": ["-y", "@stubbedev/ms-graph-dev-mcp"]
    }
  }
}

Windsurf

Edit ~/.codeium/windsurf/mcp_config.json:

{
  "mcpServers": {
    "ms-graph-dev": {
      "command": "npx",
      "args": ["-y", "@stubbedev/ms-graph-dev-mcp"]
    }
  }
}

Zed

Edit your settings.json (open via Zed → Settings → Open Settings):

{
  "context_servers": {
    "ms-graph-dev": {
      "command": {
        "path": "npx",
        "args": ["-y", "@stubbedev/ms-graph-dev-mcp"]
      }
    }
  }
}

OpenCode

Edit ~/.config/opencode/config.json:

{
  "mcp": {
    "ms-graph-dev": {
      "type": "local",
      "command": ["npx", "-y", "@stubbedev/ms-graph-dev-mcp"]
    }
  }
}

Codex (OpenAI)

Edit ~/.codex/config.json:

{
  "mcpServers": {
    "ms-graph-dev": {
      "command": "npx",
      "args": ["-y", "@stubbedev/ms-graph-dev-mcp"]
    }
  }
}

Development

git clone https://github.com/stubbedev/ms-graph-dev-mcp.git
cd ms-graph-dev-mcp
npm install
npm run build
npm start

For live reload during development:

npm run dev

Test with MCP Inspector

npx @modelcontextprotocol/inspector npx -y @stubbedev/ms-graph-dev-mcp

License

MIT

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
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
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
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
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
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
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
mcp-server-qdrant

mcp-server-qdrant

这个仓库展示了如何为向量搜索引擎 Qdrant 创建一个 MCP (Managed Control Plane) 服务器的示例。

官方
精选