Skip to content
/ MIST Public

MCP server empowering AI assistants with real-world capabilities: Gmail, Calendar, Tasks, Git integration, and note management. Bridges AI assistants to external services through standardized protocol with secure authentication.

License

Notifications You must be signed in to change notification settings

CLoaKY233/MIST

Repository files navigation

MIST - Model Intelligence System for Tasks

MIST Banner

MIST is a comprehensive MCP (Model Context Protocol) server that empowers AI assistants with real-world capabilities for note management, Gmail, Calendar, Tasks, and Git integration. It bridges the gap between AI assistants and external tools, enabling them to interact with your data and services through a standardized protocol.

License: MIT Python Version Version

Overview

MIST allows AI assistants to:

  • Manage notes: Create, edit, search, and organize notes with persistent storage
  • Interact with Gmail: Search emails, read messages, send emails, manage labels
  • Manage calendar events: View, create, update, and delete Google Calendar events
  • Organize tasks: Create task lists, add tasks, mark tasks as complete
  • Work with Git repositories: Perform Git operations like status, diff, commit, branch management

By implementing the Model Context Protocol (MCP), MIST enables AI assistants to perform actions in external systems, access real-time information, and maintain context across interactions.

Key Features

Note Management

  • Create and store notes with rich text formatting
  • Search notes by content, tags, or metadata
  • Generate summaries of notes or collections
  • Organize notes by subject, date, or custom categories
  • Store notes locally in your filesystem

Gmail Integration

  • Search emails with advanced filtering options
  • Read complete email content
  • Compose and send emails
  • Manage email labels and organization
  • Mark messages as read/unread
  • Batch operations for multiple messages

Calendar Integration

  • List all available calendars
  • Create and manage calendar events
  • Update existing events
  • Search events by keyword or date range
  • View detailed event information

Tasks Integration

  • List and manage task lists
  • Create tasks with titles, notes, and due dates
  • Mark tasks as complete
  • Delete completed tasks
  • Organize tasks into different lists

Git Integration

  • Perform Git operations in local repositories
  • View status, diffs, and commit history
  • Commit changes and manage branches
  • Create, checkout and merge branches
  • Manage remote repositories
  • Display commit details and changes

Future Roadmap

MIST is continuously evolving to offer more integrations and features. Here's what we're planning to add:

Document & File Management

  • Google Drive Integration: Access and manage files, search documents, and analyze content
  • Google Sheets Integration: Create, read, and update spreadsheets for data analysis

Search & Information Access

  • Google Custom Search: Enable web search capabilities for research and information retrieval
  • Knowledge Base: Build local repositories of information for offline access

Collaboration Tools

  • Google Classroom: Access courses, assignments, and educational materials
  • Google Chat: Interact with chat spaces and messages for team collaboration

Data & Location Services

  • Google Forms: Create and manage forms for data collection
  • Google Maps: Add location-based services including geocoding and directions

Platform Improvements

  • Enhanced Caching: Improve performance for frequently accessed data
  • Multi-User Support: Add user profiles and permission management
  • Expanded AI Assistant Compatibility: Support more AI assistants and models

We welcome contributions in any of these areas! Check our Contributing Guidelines for more information on how to get involved.

System Architecture

MIST operates as an MCP server that connects AI assistants to various services. It handles authentication, data retrieval, and actions with the appropriate services, returning structured data back to the AI assistant.

flowchart TB
    User[User] -->|Requests action| AI[AI Assistant]
    AI -->|Sends MCP request| MIST[MIST Server]

    subgraph "MIST Server Components"
        Router[MCP Request Router]
        NoteMgr[Note Manager]
        GmailMgr[Gmail Manager]
        TasksMgr[Tasks Manager]
        CalendarMgr[Calendar Manager]
        GitMgr[Git Manager]
        Auth[Authentication]
    end

    MIST --> Router
    Router -->|Note requests| NoteMgr
    Router -->|Gmail requests| GmailMgr
    Router -->|Tasks requests| TasksMgr
    Router -->|Calendar requests| CalendarMgr
    Router -->|Git requests| GitMgr
    GmailMgr -->|Auth requests| Auth
    TasksMgr -->|Auth requests| Auth
    CalendarMgr -->|Auth requests| Auth

    NoteMgr -->|Store/Retrieve| NoteDB[(Note Storage)]
    GmailMgr -->|API Calls| Gmail[(Gmail API)]
    TasksMgr -->|API Calls| Tasks[(Tasks API)]
    CalendarMgr -->|API Calls| Calendar[(Calendar API)]
    GitMgr -->|Git Operations| GitRepos[(Git Repositories)]

    MIST -->|Returns response| AI
    AI -->|Presents result| User
Loading

Communication Flow

sequenceDiagram
    participant User
    participant AI as AI Assistant
    participant MIST as MIST Server
    participant Services as External Services

    User->>AI: "Send an email to John"
    AI->>MIST: MCP Request (action: sendEmail)
    MIST->>Services: Authenticate & API Call
    Services->>MIST: Response
    MIST->>AI: Structured Response
    AI->>User: "Email sent to John"

    User->>AI: "Take notes about my project"
    AI->>MIST: MCP Request (action: createNote)
    MIST->>Services: Store Note
    Services->>MIST: Confirmation
    MIST->>AI: Success Response
    AI->>User: "Notes saved successfully"
Loading

Installation

Prerequisites

  • Python 3.13 or newer
  • UV package manager (recommended) or pip
  • Google account (for Gmail/Calendar/Tasks integration)
  • Git (for Git integration features)

Setup

  1. Clone this repository:

    git clone https://github.com/cloaky233/mist.git
    cd mist
  2. Install dependencies:

    With UV (recommended):

    uv install

    With pip:

    pip install -e .
  3. Configure your environment:

    # Create a .env file with your configuration
    echo "MIST_NOTES_DIR=/path/to/your/notes/directory" > .env
    echo "MIST_GOOGLE_CREDENTIALS_PATH=./credentials.json" >> .env
    echo "MIST_GOOGLE_TOKEN_PATH=./token.json" >> .env

    Edit the .env file to set the correct paths for your environment.

Environment Configuration

The following environment variables can be set in your .env file:

# Notes storage configuration
MIST_NOTES_DIR=/path/to/notes/directory

# Google API configuration
MIST_GOOGLE_CREDENTIALS_PATH=./credentials.json
MIST_GOOGLE_TOKEN_PATH=./token.json

Google API Configuration

Setting up Gmail, Calendar, and Tasks APIs

flowchart TD
    A[Start] --> B[Create Google Cloud Project]
    B --> C[Enable Required APIs]
    C --> D[Configure OAuth Consent Screen]
    D --> E[Create OAuth Credentials]
    E --> F[Download credentials.json]
    F --> G[Run MIST First Time]
    G --> H[Complete OAuth Flow]
    H --> I[OAuth Token Saved]
    I --> J[MIST Ready for Google Integration]
Loading
  1. Create a Google Cloud Project

    • Navigate to the Google Cloud Console
    • Create a new project or select an existing one
    • Give it a descriptive name (e.g., "MIST Integration")
  2. Enable Required APIs

    • Go to "APIs & Services" > "Library"
    • Search for and enable:
      • Gmail API
      • Calendar API
      • Tasks API
  3. Configure OAuth Consent Screen

    • Go to "APIs & Services" > "OAuth consent screen"
    • Select "External" user type (or "Internal" for Google Workspace)
    • Fill in required information:
      • App name: "MIST Integration"
      • User support email: Your email address
      • Developer contact information: Your email address
    • Click "Save and Continue"
    • Add necessary scopes for each API
    • Add your Google account as a test user
  4. Create OAuth Credentials

    • Go to "APIs & Services" > "Credentials"
    • Click "Create Credentials" > "OAuth client ID"
    • Select "Desktop application" as the application type
    • Enter a name (e.g., "MIST Desktop Client")
    • Download the JSON file and save it as credentials.json in your project directory

Usage

Running the MCP Server

Start the MCP server with:

uv run --with mcp[cli] --with-editable . mcp run mist/server.py

For production use, you may want to use a process manager like systemd, supervisor, or pm2.

Connecting to AI Assistants

Claude Desktop

Add the following to your Claude Desktop settings:

{
  "mcpServers": {
    "M.I.S.T.": {
      "command": "uv",
      "args": [
        "run",
        "--with",
        "mcp[cli]",
        "--with-editable",
        "/path/to/mist",
        "mcp",
        "run",
        "/path/to/mist/server.py"
      ],
      "env": {
        "MIST_GOOGLE_CREDENTIALS_PATH": "/path/to/mist/credentials.json",
        "MIST_GOOGLE_TOKEN_PATH": "/path/to/mist/token.json",
        "MIST_NOTES_DIR": "/path/to/notes"
      }
    }
  }
}

Zed Editor

Add the following to your Zed editor settings:

"context_servers": {
  "mist-server": {
    "command": {
      "path": "uv",
      "args": [
        "run",
        "--with",
        "mcp[cli]",
        "--with-editable",
        "/path/to/mist",
        "mcp",
        "run",
        "/path/to/mist/server.py"
      ],
      "env": {
        "MIST_GOOGLE_CREDENTIALS_PATH": "/path/to/mist/credentials.json",
        "MIST_GOOGLE_TOKEN_PATH": "/path/to/mist/token.json",
        "MIST_NOTES_DIR": "/path/to/notes"
      }
    },
    "settings": {
      "enable_server": true
    }
  }
}

First Run Authentication

When you first run the server, you'll need to authenticate with Google:

  1. A browser window will open with Google's authentication page
  2. Sign in with your Google account
  3. Grant the requested permissions
  4. The authorization code will be captured automatically
  5. MIST will save the authentication token for future use

API Reference

Note Management API

Function Description Parameters
add_note Create a new note title, content, subject (optional)
read_note Read a note by ID or title note_id or title
list_notes List notes with filters subject, tag, limit
search_notes Search notes by content query
edit_note Update an existing note note_id, new_content
delete_note Delete a note note_id

Gmail API

Function Description Parameters
search_emails Find emails by criteria Various filters
get_emails Get email content by IDs message_ids
send_email Send a new email to, subject, body, etc.
list_available_labels Get Gmail labels None

Calendar API

Function Description Parameters
list_calendars_tool List all calendars None
create_event_tool Create a calendar event calendar_id, title, etc.
search_events_tool Search calendar events calendar_id, query, etc.

Tasks API

Function Description Parameters
list_task_lists_tool List all task lists None
create_task_tool Create a new task task_list_id, title, etc.
complete_task_tool Mark task as complete task_list_id, task_id

Git API

Below is a selection of commonly used Git tools. For a complete list and detailed parameters, please refer to the API Reference.

Function Description Key Parameters
status_tool Show working tree status repo_path
diff_unstaged_tool Show unstaged changes repo_path
diff_staged_tool Show staged changes repo_path
add_tool Stage files for commit repo_path, files
reset_tool Reset current HEAD or unstage files repo_path, commit_ish (opt), mode (opt)
commit_tool Commit staged changes repo_path, message
log_tool Show commit history repo_path, max_count (opt)
branch_list_tool List all branches repo_path
create_branch_tool Create a new branch repo_path, branch_name, base_branch (opt)
checkout_tool Switch branches repo_path, branch_name
merge_tool Merge a branch or manage merge state repo_path, branch (opt), abort (opt)
rebase_tool Rebase current branch or manage rebase state repo_path, branch_or_commit (opt), interactive (opt), abort (opt), continue_rebase (opt)
stash_save_tool Stash current changes repo_path, message (opt)
pull_tool Pull changes from a remote repo_path, remote (opt), branch (opt)
push_tool Push changes to a remote repo_path, remote (opt), branch (opt), force (opt), tags (opt)

Project Structure

mist/
├── __init__.py               # Package initialization with version info
├── server.py                 # Main MCP server entry point
├── tools/                    # Tool implementations
│   ├── note_tools/           # Note management
│   │   ├── __init__.py
│   │   └── tool.py           # Note tool implementations
│   ├── gmail_tools/          # Gmail integration
│   │   ├── __init__.py
│   │   ├── config.py         # Gmail-specific config
│   │   ├── gmail.py          # Gmail API utilities
│   │   └── tool.py           # Gmail tool implementations
│   ├── calendar_tools/       # Calendar integration
│   │   ├── __init__.py
│   │   ├── calendar.py       # Calendar API utilities
│   │   ├── config.py         # Calendar-specific config
│   │   └── tool.py           # Calendar tool implementations
│   ├── tasks_tools/          # Tasks integration
│   │   ├── __init__.py
│   │   ├── config.py         # Tasks-specific config
│   │   ├── tasks.py          # Tasks API utilities
│   │   └── tool.py           # Tasks tool implementations
│   ├── git_tools/            # Git operations
│   │   ├── __init__.py
│   │   ├── config.py         # Git-specific config (imports from google_api)
│   │   ├── gitapi.py         # Core Git command implementations
│   │   └── tool.py           # Git tool functions exposed to MCP
│   ├── google_api/           # Common Google API utilities
│   │   ├── __init__.py
│   │   ├── client.py         # Google API client
│   │   └── config.py         # Shared Google API config
│   └── __init__.py           # Tools package init
├── config_docs/              # Configuration documentation
│   ├── api_reference.md      # API reference documentation
│   ├── configs.md            # Example configuration templates
│   ├── configuration_guide.md# Detailed configuration guide
│   └── troubleshooting_guide.md # Troubleshooting guide
├── .github/                  # GitHub configuration
│   └── ISSUE_TEMPLATE/       # Issue templates for bug reports, etc.
├── LICENSE                   # MIT License
├── CONTRIBUTING.md           # Contributing guidelines
├── README.md                 # Project documentation
├── pyproject.toml            # Project metadata and dependencies
├── uv.lock                   # UV lock file for dependencies
├── credentials.json          # Google API credentials (should be user-provided)
└── .env                      # Environment configuration (user-created)

Security Considerations

  • API keys and tokens are stored in the .env file or environment variables
  • OAuth tokens are securely managed by Google's authentication flow
  • Your data remains on your own device or within your own accounts
  • No data is sent to third-party services unless explicitly configured

Troubleshooting

Authentication Issues

  • Credentials not working: Delete token.json and re-authenticate
  • Permission errors: Check OAuth consent screen for proper scopes
  • Token expired: MIST should refresh tokens automatically, but you may need to re-authenticate if refresh fails

Connection Problems

  • AI assistant cannot connect: Verify paths in configuration are correct
  • Server fails to start: Check Python version (3.13+ required)
  • API errors: Verify API keys and credentials are correctly set up

Data Issues

  • Notes not saving: Check MIST_NOTES_DIR path exists and is writable

Development

Linting and Formatting

ruff check .
ruff format .

Contributing

Contributions are welcome! Please see our Contributing Guidelines for details on how to get started.

  1. Fork the repository
  2. Create a feature branch: git checkout -b feature/amazing-feature
  3. Make your changes and commit them: git commit -m 'Add amazing feature'
  4. Push to your branch: git push origin feature/amazing-feature
  5. Open a pull request

License

This project is licensed under the MIT License - see the LICENSE file for details.

Acknowledgments

  • Model Context Protocol for the protocol specification
  • Google APIs for Gmail, Calendar, and Tasks integration

Made with ❤️ by CLoaKY

About

MCP server empowering AI assistants with real-world capabilities: Gmail, Calendar, Tasks, Git integration, and note management. Bridges AI assistants to external services through standardized protocol with secure authentication.

Topics

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages