Teams MCP Server
Enables AI assistants to interact with Microsoft Teams through direct API calls, supporting message search, sending replies, managing favorites, and more without Azure AD registration.
README
Teams MCP Server
An MCP (Model Context Protocol) server that enables AI assistants to interact with Microsoft Teams. Search messages, send replies, manage favourites, and more.
How It Works
This server calls Microsoft's Teams APIs directly (Substrate, chatsvc, CSA) - the same APIs the Teams web app uses. No Azure AD app registration or admin consent required.
Authentication flow:
- AI runs
teams_loginto open a browser for you to log in - OAuth tokens are extracted and cached
- All operations use cached tokens directly (no browser needed)
- Automatic token refresh (~1 hour)
Security: Uses the same authentication as the Teams web client - your access is limited to what your account can already do.
Installation
Prerequisites
- Node.js 18+
- A Microsoft account with Teams access
- Google Chrome, Microsoft Edge, or Chromium browser installed
Configure Your MCP Client
Add to your MCP client configuration (e.g., Claude Desktop, Windsurf, Cursor):
{
"mcpServers": {
"teams": {
"command": "npx",
"args": ["-y", "msteams-mcp@latest"]
}
}
}
That's it. npx will automatically download and run the latest version.
From Source (alternative)
If you prefer to run from a local clone:
git clone https://github.com/m0nkmaster/msteams-mcp.git
cd msteams-mcp
npm install && npm run build
Then configure your MCP client:
{
"mcpServers": {
"teams": {
"command": "node",
"args": ["/path/to/msteams-mcp/dist/index.js"]
}
}
}
The server uses your system's Chrome (macOS/Linux) or Edge (Windows) for authentication.
Available Tools
Search & Discovery
| Tool | Description |
|---|---|
teams_search |
Search Teams messages with operators (from:, sent:, in:, hasattachment:, etc.) |
teams_search_email |
Search emails in your mailbox (same auth as Teams — no extra login) |
teams_get_message |
Get a single message by ID with full content (any age) |
teams_get_thread |
Get messages from a conversation/thread |
teams_find_channel |
Find channels by name (your teams + org-wide discovery) |
teams_get_activity |
Get activity feed (mentions, reactions, replies, notifications) |
Messaging
| Tool | Description |
|---|---|
teams_send_message |
Send a message (default: self-chat/notes). Use replyToMessageId for thread replies |
teams_edit_message |
Edit one of your own messages |
teams_delete_message |
Delete one of your own messages (soft delete) |
People & Contacts
| Tool | Description |
|---|---|
teams_get_me |
Get current user profile (email, name, ID) |
teams_search_people |
Search for people by name or email |
teams_get_frequent_contacts |
Get frequently contacted people (useful for name resolution) |
teams_get_chat |
Get conversation ID for 1:1 chat with a person |
teams_create_group_chat |
Create a new group chat with multiple people (2+ others) |
Organisation
| Tool | Description |
|---|---|
teams_get_favorites |
Get pinned/favourite conversations |
teams_add_favorite |
Pin a conversation |
teams_remove_favorite |
Unpin a conversation |
teams_save_message |
Bookmark a message |
teams_unsave_message |
Remove bookmark from a message |
teams_get_saved_messages |
Get list of saved/bookmarked messages with source references |
teams_get_followed_threads |
Get list of followed threads with source references |
teams_get_unread |
Get unread counts (aggregate or per-conversation) |
teams_mark_read |
Mark a conversation as read up to a message |
Reactions
| Tool | Description |
|---|---|
teams_search_emoji |
Search for emojis by name (standard + custom org emojis) |
teams_add_reaction |
Add an emoji reaction to a message |
teams_remove_reaction |
Remove an emoji reaction from a message |
Quick reactions: like, heart, laugh, surprised, sad, angry can be used directly without searching.
Calendar & Meetings
| Tool | Description |
|---|---|
teams_get_meetings |
Get meetings from calendar (defaults to next 7 days) |
teams_get_transcript |
Get meeting transcript (requires threadId from teams_get_meetings) |
teams_get_meetings returns: subject, times, organiser, join URL, threadId for meeting chat. Use threadId with teams_get_thread to read meeting chat, or with teams_get_transcript to get the full transcript with speakers and timestamps.
Files
| Tool | Description |
|---|---|
teams_get_shared_files |
Get files and links shared in a conversation (supports pagination) |
Returns both files (name, extension, URL, size) and links (URL, title), along with who shared each item. Works for channels, group chats, 1:1 chats, and meeting chats.
Session
| Tool | Description |
|---|---|
teams_login |
Trigger manual login (opens browser) |
teams_status |
Check authentication and session state |
Search Operators
Both teams_search (Teams messages) and teams_search_email (emails) support native operators:
from:sarah@company.com # Messages/emails from person
sent:2026-01-20 # From specific date
sent:>=2026-01-15 # Since date
in:project-alpha # Messages in channel (Teams only)
subject:"budget" # By subject (email)
"Rob Smith" # Find @mentions (name in quotes)
hasattachment:true # With files
is:unread # Unread emails (email only)
NOT from:email@co.com # Exclude results
Combine operators: from:sarah@co.com sent:>=2026-01-18 hasattachment:true
Note: @me, from:me, to:me do NOT work. Use teams_get_me first to get your email/displayName. sent:today works, but sent:lastweek and sent:thisweek do NOT - use explicit dates or omit (results are sorted by recency).
MCP Resources
The server also exposes passive resources for context discovery:
| Resource URI | Description |
|---|---|
teams://me/profile |
Current user's profile |
teams://me/favorites |
Pinned conversations |
teams://status |
Authentication status |
CLI Tools (Development)
For local development, CLI tools are available for testing and debugging:
# Check authentication status
npm run cli -- status
# Search messages
npm run cli -- search "meeting notes"
npm run cli -- search "project" --from 0 --size 50
# Search emails
npm run cli -- teams_search_email --query "from:sarah@company.com"
# Send messages (default: your own notes/self-chat)
npm run cli -- send "Hello from Teams MCP!"
npm run cli -- send "Message" --to "conversation-id"
# Force login
npm run cli -- login --force
# Output as JSON
npm run cli -- search "query" --json
MCP Test Harness
Test the server through the actual MCP protocol:
# List available tools
npm run cli
# Call any tool
npm run cli -- search "your query"
npm run cli -- status
npm run cli -- people "john smith"
npm run cli -- favorites
npm run cli -- activity # Get activity feed
npm run cli -- unread # Check unread counts
npm run cli -- teams_search_emoji --query "heart" # Search emojis
Limitations
- Login required - Run
teams_loginto authenticate (opens browser) - Token expiry - Tokens expire after ~1 hour; headless refresh is attempted or run
teams_loginagain when needed - Undocumented APIs - Uses Microsoft's internal APIs which may change without notice
- Search limitations - Full-text search only; thread replies not matching search terms won't appear (use
teams_get_threadfor full context) - Own messages only - Edit/delete only works on your own messages
Session Files
Session files are stored in a user config directory (encrypted):
- macOS/Linux:
~/.teams-mcp-server/ - Windows:
%APPDATA%\teams-mcp-server\
Contents: session-state.json, token-cache.json, browser-profile/
If your session expires, call teams_login or delete the config directory.
Development
For local development:
git clone https://github.com/m0nkmaster/msteams-mcp.git
cd msteams-mcp
npm install
npm run build
Development commands:
npm run dev # Run MCP server in dev mode
npm run build # Compile TypeScript
npm run lint # Run ESLint
npm run research # Explore Teams APIs (logs network calls)
npm test # Run unit tests
npm run typecheck # TypeScript type checking
For development with hot reload, configure your MCP client:
{
"mcpServers": {
"teams": {
"command": "npx",
"args": ["tsx", "/path/to/msteams-mcp/src/index.ts"]
}
}
}
See AGENTS.md for detailed architecture and contribution guidelines.
Teams Chat Export Bookmarklet
This repo also includes a standalone bookmarklet for exporting Teams chat messages to Markdown. See teams-bookmarklet/README.md.
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.