
Microsoft MCP
A powerful MCP server that enables AI assistants to interact with Microsoft Graph API for managing Outlook emails, Calendar events, OneDrive files, and Contacts through natural language commands.
Tools
list_accounts
List all signed-in Microsoft accounts
read_emails
Read emails from any folder (set include_body=False for subject/sender only)
get_email
Get full email details including attachments list
reply_to_email
Reply to an email maintaining thread context
mark_email_read
Mark email as read or unread
move_email
Move email to another folder (drafts, sentitems, deleteditems, etc)
download_attachment
Download email attachment as base64
send_email
Send email with optional CC and attachments (attachments = [{"name": "file.txt", "content_base64": "..."}])
get_calendar_events
Get calendar events with full details
check_availability
Check free/busy availability for scheduling
create_event
Create calendar event with full details
update_event
Update existing calendar event
delete_event
Delete/cancel calendar event
respond_to_event
Respond to meeting invitation (accept, decline, tentativelyAccept)
get_contacts
Get contacts/people with email addresses
list_files
List files with pagination support (returns items and @odata.nextLink)
download_file
Download file content as base64
upload_file
Upload file to OneDrive
delete_file
Delete file or folder
search
Search across emails, files, events, and people
README
Microsoft MCP
Powerful MCP server for Microsoft Graph API - a complete AI assistant toolkit for Outlook, Calendar, OneDrive, and Contacts.
Features
- Email Management: Read, send, reply, manage attachments, organize folders
- Calendar Intelligence: Create, update, check availability, respond to invitations
- OneDrive Files: Upload, download, browse with pagination
- Contacts: Search and list contacts from your address book
- Multi-Account: Support for multiple Microsoft accounts (personal, work, school)
- Unified Search: Search across emails, files, events, and people
Quick Start with Claude Desktop
# Add Microsoft MCP server (replace with your Azure app ID)
claude mcp add microsoft-mcp -e MICROSOFT_MCP_CLIENT_ID=your-app-id-here -- uvx --from git+https://github.com/elyxlz/microsoft-mcp.git microsoft-mcp
# Start Claude Desktop
claude
Usage Examples
# Email examples
> read my latest emails with full content
> reply to the email from John saying "I'll review this today"
> send an email with attachment to alice@example.com
# Calendar examples
> show my calendar for next week
> check if I'm free tomorrow at 2pm
> create a meeting with Bob next Monday at 10am
# File examples
> list files in my OneDrive
> upload this report to OneDrive
> search for "project proposal" across all my files
# Multi-account
> list all my Microsoft accounts
> send email from my work account
Available Tools
Email Tools
read_emails
- Read emails with full body contentget_email
- Get specific email with attachmentsreply_to_email
- Reply maintaining thread contextsend_email
- Send with CC/BCC and attachmentsmark_email_read
- Mark emails as read/unreadmove_email
- Move between foldersdownload_attachment
- Download email attachments
Calendar Tools
get_calendar_events
- List events with full detailscheck_availability
- Check free/busy timescreate_event
- Create with location and attendeesupdate_event
- Reschedule or modify eventsdelete_event
- Cancel eventsrespond_to_event
- Accept/decline invitations
Contact & File Tools
get_contacts
- Search or list contactslist_files
- Browse OneDrive with paginationupload_file
- Upload files to OneDrivedownload_file
- Download file contentdelete_file
- Delete files or folders
Utility Tools
search
- Universal search across all serviceslist_accounts
- Show authenticated accounts
Manual Setup
1. Azure App Registration
- Go to Azure Portal → Microsoft Entra ID → App registrations
- New registration → Name:
microsoft-mcp
- Supported account types: Personal + Work/School
- Authentication → Allow public client flows: Yes
- API permissions → Add these delegated permissions:
- Mail.ReadWrite
- Calendars.ReadWrite
- Files.ReadWrite
- Contacts.Read
- People.Read
- User.Read
- Copy Application ID
2. Installation
git clone https://github.com/elyxlz/microsoft-mcp.git
cd microsoft-mcp
uv sync
3. Authentication
# Set your Azure app ID
export MICROSOFT_MCP_CLIENT_ID="your-app-id-here"
# Run authentication script
uv run authenticate.py
# Follow the prompts to authenticate your Microsoft accounts
4. Claude Desktop Configuration
Add to your Claude Desktop configuration:
macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%\Claude\claude_desktop_config.json
{
"mcpServers": {
"microsoft": {
"command": "uvx",
"args": ["--from", "git+https://github.com/elyxlz/microsoft-mcp.git", "microsoft-mcp"],
"env": {
"MICROSOFT_MCP_CLIENT_ID": "your-app-id-here"
}
}
}
}
Or for local development:
{
"mcpServers": {
"microsoft": {
"command": "uv",
"args": ["--directory", "/path/to/microsoft-mcp", "run", "microsoft-mcp"],
"env": {
"MICROSOFT_MCP_CLIENT_ID": "your-app-id-here"
}
}
}
}
Multi-Account Support
All tools support an optional account_id
parameter:
# Use default account
send_email("user@example.com", "Subject", "Body")
# Use specific account
send_email("user@example.com", "Subject", "Body", account_id="account-id-here")
# List accounts to get IDs
list_accounts()
Development
# Run tests
uv run pytest tests/ -v
# Type checking
uv run pyright
# Format code
uvx ruff format .
# Lint
uvx ruff check --fix --unsafe-fixes .
Example: AI Assistant Scenarios
Smart Email Management
# Read latest emails with full content
emails = read_emails(count=10, include_body=True)
# Reply maintaining thread
reply_to_email(email_id, "Thanks for your message. I'll review and get back to you.")
# Forward with attachments
email = get_email(email_id)
attachments = [download_attachment(email_id, att["id"]) for att in email["attachments"]]
send_email("boss@company.com", f"FW: {email['subject']}", email["body"]["content"], attachments=attachments)
Intelligent Scheduling
# Check availability before scheduling
availability = check_availability("2024-01-15T10:00:00Z", "2024-01-15T18:00:00Z", ["colleague@company.com"])
# Create meeting with details
create_event(
"Project Review",
"2024-01-15T14:00:00Z",
"2024-01-15T15:00:00Z",
location="Conference Room A",
body="Quarterly review of project progress",
attendees=["colleague@company.com", "manager@company.com"]
)
Security Notes
- Tokens are cached locally in
~/.microsoft_mcp_token_cache.json
- Use app-specific passwords if you have 2FA enabled
- Only request permissions your app actually needs
- Consider using a dedicated app registration for production
Troubleshooting
- Authentication fails: Check your CLIENT_ID is correct
- "Need admin approval": Use
MICROSOFT_MCP_TENANT_ID=consumers
for personal accounts - Missing permissions: Ensure all required API permissions are granted in Azure
- Token errors: Delete
~/.microsoft_mcp_token_cache.json
and re-authenticate
License
MIT
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.

VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.

E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.