Frequently Asked Questions (FAQ) about Teams Messenger MCP App

Q: What is the Model Context Protocol (MCP)? A: MCP is an open protocol standardizing how applications provide context to Large Language Models (LLMs), enabling AI models to access and interact with external data sources and tools in a consistent manner.

Q: What does the Teams Messenger MCP App do? A: It bridges Microsoft Teams with MCP-compatible clients (LLMs, agentic frameworks) using a pure MCP server, allowing AI functionalities within Teams without requiring REST API endpoints.

Q: What are the key features of this MCP App? A: Key features include Microsoft Teams chat integration, PostgreSQL-based Information Retrieval (IR) server, persistent storage in DuckDB, hybrid semantic/lexical search, a CLI for token management, and live event streaming for messages.

Q: What is the architecture of the Teams Messenger MCP App? A: It connects Microsoft Teams to an MCP Server (Python, FastMCP), which interacts with a DuckDB database for storage and an IR Server (PostgreSQL) for advanced search.

Q: What are the installation requirements? A: Requirements include Python 3.9+, pip, and Docker/Docker Compose for containerized deployment. Local installation is also possible.

Q: What environment variables need to be configured? A: Required variables include Azure AD credentials (client ID, secret, tenant ID, app object ID), database paths (DuckDB and token), polling interval, and PostgreSQL/IR server settings.

Q: How do I run the MCP server locally? A: After installing dependencies and configuring environment variables, run python mcp_server/server.py.

Q: How do I use the CLI client? A: Use the CLI to login, list chats, get messages, send messages, create chats, search messages (hybrid, BM25, vector), and stream new messages.

Q: What search capabilities does the IR server provide? A: The IR server offers advanced search using PostgreSQL and pgvector, including hybrid (BM25 + vector) search with LLM reranking.

Q: Can I test the app without real Teams API calls? A: Yes, set DEMO_MODE=true in your .env file to use mock/demo mode.

Q: What is UBOS and how does it enhance this MCP App? A: UBOS is a full-stack AI Agent Development Platform that helps you orchestrate AI Agents, connect them with enterprise data, build custom AI Agents with your LLM model and Multi-Agent Systems. It provides a more comprehensive environment for AI agent development and management compared to the standalone MCP App.