Frequently Asked Questions (FAQ)

Q: What is the Teams Messenger MCP App? A: The Teams Messenger MCP App is a pure Model Context Protocol (MCP) server that bridges Microsoft Teams and MCP-compatible clients (LLMs, agentic frameworks, and a rich CLI MCP client). It exposes all features via MCP tools, resources, and events—no REST API endpoints.

Q: What are the key features of the Teams Messenger MCP App? A: Key features include Microsoft Teams chat/message integration via MCP, a PostgreSQL-based Information Retrieval (IR) server for advanced search, persistent storage in DuckDB for chat/message history, hybrid semantic and lexical search, CLI for login/token management, polling-based event emission for new messages, live event streaming and search, and single-agent (bot) account support.

Q: What is MCP (Model Context Protocol)? A: MCP is an open protocol that standardizes how applications provide context to LLMs. It allows AI models to access and interact with external data sources and tools.

Q: How does the Teams Messenger MCP App integrate with Microsoft Teams? A: The app integrates with Microsoft Teams via the Graph API, enabling seamless communication and data exchange between Teams and AI agents.

Q: What is the purpose of the IR server in the Teams Messenger MCP App? A: The IR server provides advanced search capabilities within Teams data. It is powered by PostgreSQL and pgvector and supports hybrid semantic and lexical search.

Q: What is DuckDB used for in the Teams Messenger MCP App? A: DuckDB is used for persistent storage of chat and message history, ensuring data persistence and facilitating historical analysis.

Q: What is hybrid semantic and lexical search? A: Hybrid semantic and lexical search combines semantic understanding with lexical precision, delivering highly relevant search results by considering both the meaning of the query and the keywords used.

Q: How do I install the Teams Messenger MCP App? A: You can install the app locally or via Docker. Local installation involves cloning the repository, installing dependencies, and configuring environment variables. Docker deployment simplifies the installation process and ensures consistent performance.

Q: What environment variables do I need to configure for the Teams Messenger MCP App? A: You need to configure variables such as AZURE_CLIENT_ID, AZURE_CLIENT_SECRET, AZURE_TENANT_ID, DUCKDB_PATH, TOKEN_PATH, OPENAI_API_KEY, POSTGRES_USER, and others, as detailed in the .env.template file.

Q: How do I run the MCP server? A: You can run the MCP server in local mode (using a Python script) or in Docker mode (using Docker Compose).

Q: What is demo mode in the Teams Messenger MCP App? A: Demo mode simulates Teams API calls without interacting with the actual Teams environment, allowing you to test and demonstrate the app without real Teams data.

Q: How do I use the CLI to interact with the MCP server? A: The CLI provides commands for listing chats, retrieving messages, sending messages, creating new chats, and searching messages. Use commands like list_chats, get_messages, send_message, create_chat, and search_messages.

Q: What are the API endpoints for the IR server? A: The IR server provides endpoints for health checks (GET /), listing available tools (GET /api/tools), searching content (POST /api/tools/search), and indexing content (POST /api/tools/index_content).

Q: How do I troubleshoot common issues with the Teams Messenger MCP App? A: Check your Azure AD credentials, ensure polling is running, verify file permissions and paths, and review Docker logs for any errors. Set DEMO_MODE=true for local testing without real Teams.

Q: How does the UBOS platform enhance the Teams Messenger MCP App? A: The UBOS platform provides AI agent orchestration, enterprise data connectivity, custom AI agent building, and multi-agent systems, enabling you to unlock the full potential of the Teams Messenger MCP App.