概览
What is Introduction?
Introduction is a Google Chat integration for MCP (Model Control Protocol) servers, written in Python with FastMCP. It provides tools to access and interact with Google Chat spaces and messages, and includes a separate authentication server for OAuth2 login. It is intended for developers who want to connect AI assistants or other MCP clients to Google Chat.
How to use Introduction?
Clone the repository, set up a Google Cloud project with the Chat and People APIs enabled, and download OAuth2 credentials as credentials.json. Then choose an authentication mode: run uv run python server.py --auth cli for headless environments (paste a redirect URL), or uv run python server.py --auth web --port 8000 for environments with a local browser. After obtaining a token, configure your MCP client with a uv command pointing to server.py and the token path, or use the provided Docker/Podman container.
Key features of Introduction
- OAuth2 authentication with Google Chat API
- List available Google Chat spaces
- Retrieve messages from spaces with date filtering
- Local authentication server for easy setup
- Works in headless/remote environments via CLI mode
Use cases of Introduction
- Connect an AI assistant to monitor Google Chat spaces
- Automatically fetch and analyze chat history from specific spaces
- Integrate MCP‑based tools with Google Chat for workflow automation
FAQ from Introduction
What are the runtime requirements?
Python 3.13+ and a Google Cloud project with the Google Chat API and People API (for user display names) enabled.
How does authentication work?
Two modes are available: CLI mode (recommended for headless environments) displays an authorization URL that you open in any browser, then you paste the redirect URL back into the terminal. Web mode runs a local web server on port 8000 and opens the authorization page in the local browser. Both modes save a token.json file.
What tools does the server provide?
get_chat_spaces() lists all accessible spaces, and get_space_messages(space_name, start_date, end_date) retrieves messages from a specific space with optional date filtering.
Can I run Introduction in a container?
Yes. Docker and Podman images are available. Mount your project directory (containing token.json) to /data and pass --token-path=/data/token.json. For the auth server in web mode, also map port 8000.
