MS365 Email MCP Server
A Python-based MCP server for Microsoft 365 Outlook email operations using OAuth 2.0 Client Credentials Flow, enabling automated email management for autonomous agents.
README
MS365 Email MCP Server
A Python-based Model Context Protocol (MCP) server for Microsoft 365 Outlook email operations. This server uses OAuth 2.0 Client Credentials Flow for authentication, making it ideal for autonomous agents and server-to-server scenarios.
Built with FastMCP for simplified server implementation, following the pattern used by AWS API MCP Server.
Features
This MCP server provides the following email operations:
- list-mail-messages - List mail messages from inbox or a specific folder
- list-mail-folders - List all mail folders
- list-mail-folder-messages - List messages from a specific folder
- get-mail-message - Get a specific mail message by ID
- send-mail - Send an email
- delete-mail-message - Delete a mail message
- create-draft-email - Create a draft email
- move-mail-message - Move a mail message to another folder
Authentication
This server uses OAuth 2.0 Client Credentials Flow (app-only authentication), which is perfect for autonomous agents and background services. No user interaction is required.
Reference: Microsoft Graph - Get access without a user
Prerequisites
1. Azure App Registration
You need to register an application in Azure AD with Application permissions (not Delegated):
- Go to Azure Portal → Azure Active Directory → App registrations
- Create a new registration
- Note down:
- Application (client) ID
- Directory (tenant) ID
- Create a client secret:
- Go to Certificates & secrets → New client secret
- Copy the secret value immediately (it won't be shown again)
- Configure API permissions:
- Go to API permissions → Add a permission → Microsoft Graph
- Select Application permissions (not Delegated)
- Add the following permissions:
Mail.Read- Read user mailMail.ReadWrite- Read and write user mailMail.Send- Send mail as user
- Grant admin consent (required for application permissions)
Important: Application permissions require administrator consent. The app must be granted consent before it can use these permissions.
Reference: Microsoft Graph - Configure permissions
2. Environment Variables
MS365_CLIENT_ID=your-azure-ad-client-id
MS365_CLIENT_SECRET=your-azure-ad-client-secret
MS365_TENANT_ID=your-tenant-id
# Optional: For Government Cloud
MS365_CLOUD_TYPE=gov # Options: "commercial" (default), "gov", "government", "usgov"
# Optional: For shared mailboxes (required for Client Credentials Flow)
MS365_USER_IDENTIFIER=user@domain.com # UserPrincipalName or Graph ID of the shared mailbox
Important for Shared Mailboxes: When using Client Credentials Flow (app-only authentication), you must provide MS365_USER_IDENTIFIER because /me/ endpoints don't work without a signed-in user. Use the UserPrincipalName (e.g., shared-mailbox@domain.com) or the Graph ID of the shared mailbox.
Local Development
This project uses UV for dependency management, similar to the AWS API MCP Server.
1. Install UV
# Install UV using pip
pip install uv
# Or using Homebrew (macOS)
brew install uv
# Or using the official installer
curl -LsSf https://astral.sh/uv/install.sh | sh
2. Install Dependencies
# Install project dependencies
uv sync
# Or install in development mode with dev dependencies
uv sync --dev
3. Set Environment Variables
export MS365_CLIENT_ID="your-client-id"
export MS365_CLIENT_SECRET="your-client-secret"
export MS365_TENANT_ID="your-tenant-id"
export PORT="8100" # Optional, defaults to 8100
export HOST="0.0.0.0" # Optional, defaults to 0.0.0.0
export LOG_LEVEL="INFO" # Optional, defaults to INFO
export STATELESS_HTTP="true" # Optional, defaults to true
4. Run the Server
Using UV (recommended):
# Run using UV
uv run ms365-email-mcp-server
# Or run directly with Python (after uv sync)
python -m ms365_email_mcp_server.server
The server will start on http://localhost:8100 with:
- SSE endpoint:
http://localhost:8100/message - Health check:
http://localhost:8100/health
Docker Usage
Build the Docker Image
The Dockerfile uses UV for dependency management:
docker build -t email-mcp-server .
Run the Container
docker run -d \
-p 8100:8100 \
-e MS365_CLIENT_ID="your-client-id" \
-e MS365_CLIENT_SECRET="your-client-secret" \
-e MS365_TENANT_ID="your-tenant-id" \
-e MS365_CLOUD_TYPE="commercial" \
email-mcp-server
Using Docker Compose
Create a .env file:
MS365_CLIENT_ID=your-client-id
MS365_CLIENT_SECRET=your-client-secret
MS365_TENANT_ID=your-tenant-id
MS365_CLOUD_TYPE=commercial
Then run:
docker-compose up
Or in detached mode:
docker-compose up -d
MCP Client Configuration
HTTP/SSE Endpoint
The server runs as an HTTP/SSE server. Connect to it using:
- SSE Endpoint:
http://localhost:8100/message - Health Check:
http://localhost:8100/health
Using with MCP Clients
Option 1: HTTP/SSE Transport (Recommended)
For clients that support HTTP/SSE transport, configure:
{
"mcpServers": {
"ms365-email": {
"url": "http://localhost:8100/message",
"transport": "sse"
}
}
}
Option 2: Using UV (for stdio transport)
If you want to use UV to run the server (similar to AWS API MCP Server), you can configure:
{
"mcpServers": {
"ms365-email": {
"command": "uvx",
"args": [
"ms365-email-mcp-server@latest"
],
"env": {
"MS365_CLIENT_ID": "your-client-id",
"MS365_CLIENT_SECRET": "your-client-secret",
"MS365_TENANT_ID": "your-tenant-id",
"LOG_LEVEL": "INFO"
}
}
}
}
Note: This requires the package to be published to PyPI. For local development, use Option 1 or run the server manually with uv run ms365-email-mcp-server.
API Reference
Microsoft Graph API Endpoints Used
- List Messages: GET /me/messages
- Send Mail: POST /me/sendMail
- Get Message:
GET /me/messages/{id} - Delete Message:
DELETE /me/messages/{id} - Create Draft:
POST /me/messages - Move Message:
POST /me/messages/{id}/move - List Folders:
GET /me/mailFolders
Government Cloud Support
For Azure Government Cloud, set:
export MS365_CLOUD_TYPE="gov"
This will use:
- Authority:
https://login.microsoftonline.us - Graph API:
https://graph.microsoft.us
Security Notes
- Never commit your client secrets to version control
- Use environment variables or secret management systems
- Rotate client secrets regularly
- Consider using managed identities in Azure for production
- Application permissions require administrator consent
Troubleshooting
"Failed to acquire token"
- Verify your
MS365_CLIENT_ID,MS365_CLIENT_SECRET, andMS365_TENANT_IDare correct - Ensure admin consent has been granted for the application permissions
- Check that the client secret hasn't expired
"Insufficient privileges"
- Verify that admin consent has been granted for all required permissions
- Check that you're using Application permissions (not Delegated permissions)
"Invalid tenant"
- Verify your
MS365_TENANT_IDis correct - For Government Cloud, ensure
MS365_CLOUD_TYPEis set to"gov"
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.