MIST - Model Intelligence System for Tasks

What is MIST - Model Intelligence System for Tasks?

MIST is a comprehensive MCP (Model Context Protocol) server that enables AI assistants to manage notes, interact with Gmail, Google Calendar, Google Tasks, and perform Git operations. It bridges AI assistants and external services, allowing them to access real-time data and perform actions through a standardized protocol.

How to use MIST - Model Intelligence System for Tasks?

Install dependencies with UV or pip, configure environment variables for notes directory and Google credentials, then start the server with uv run --with mcp[cli] --with-editable . mcp run mist/server.py. On first run, complete the Google OAuth flow in the browser. Connect AI assistants (e.g., Claude Desktop or Zed editor) by adding the MIST MCP server configuration.

Key features of MIST - Model Intelligence System for Tasks

• Manage notes with rich text, search, and summaries
• Search, read, and send emails via Gmail API
• Create, update, and search Google Calendar events
• Manage task lists and mark tasks complete
• Perform Git operations: status, diff, commit, branch management
• OAuth authentication for Google services

Use cases of MIST - Model Intelligence System for Tasks

• An AI assistant taking meeting notes and saving them locally
• Sending an email or creating a calendar event through a chat interface
• Organizing personal tasks by creating and completing task lists
• Automating Git repository operations like committing and branching
• Searching past emails while drafting a reply with the AI assistant

FAQ from MIST - Model Intelligence System for Tasks

What runtime dependencies does MIST require?

Python 3.13 or newer, UV package manager (or pip), a Google account for Gmail/Calendar/Tasks features, and Git for Git integration.

How does authentication with Google services work?

On first run, a browser window opens for Google OAuth. After granting permissions, MIST saves the token locally (default token.json) for subsequent use.

Where are notes stored?

Notes are stored locally in the directory specified by the MIST_NOTES_DIR environment variable.

Can MIST be used with different AI assistants?

Yes, MIST is compatible with any AI assistant that supports the Model Context Protocol, with explicit configuration examples provided for Claude Desktop and Zed editor.

What integrations are currently available?

MIST supports note management, Gmail, Google Calendar, Google Tasks, and Git integration.