AI Utilities

A directory manager for AI-powered developer tools. AI Utilities provides a unified dashboard to launch and manage a growing collection of developer utilities, with automatic update notifications from GitHub Releases.

Features

• Unified dashboard – tile-based UI showing all available tools at a glance
• One-click launch – start any tool directly from the manager; it opens in your browser automatically
• GitHub-powered updates – checks for new releases on startup and notifies you when a newer version is available
• Auto-update support – optionally download and apply updates with a single click (packaged binary only)
• uv workspaces – each tool is an isolated Python workspace with its own dependencies
• macOS ARM64 executable – ships as a self-contained binary for Apple Silicon; no Python installation required

Available Applications

App	Description
Connector Tester	Test connection configurations for Zendesk, Intercom, Confluence, ServiceNow, Salesforce, and SharePoint

Getting Started

Download the binary (macOS ARM64)

Download the latest ai-utilities-macos-arm64 binary from the Releases page:

# Download
curl -L https://github.com/latalkdesk/ai-utilities/releases/latest/download/ai-utilities-macos-arm64 -o ai-utilities

# Make executable
chmod +x ai-utilities

# Run
./ai-utilities

The manager opens at http://localhost:8000 in your default browser.

Run from source

Requirements: uv and Python 3.13+

git clone https://github.com/latalkdesk/ai-utilities
cd ai-utilities

# Install all workspace dependencies
uv sync --all-packages

# Start the manager
uv run ai-utilities

Project Structure

ai-utilities/
├── pyproject.toml              # Root workspace definition
├── uv.lock                     # Shared lockfile
├── build.py                    # PyInstaller build script
├── README.md
│
├── manager/                    # Main manager application
│   ├── pyproject.toml
│   └── src/ai_utilities_manager/
│       ├── main.py             # FastAPI dashboard app
│       ├── app_registry.py     # Workspace app discovery
│       ├── updater.py          # GitHub release checker & auto-updater
│       ├── config.py           # User settings (~/.ai-utilities/config.json)
│       └── static/             # Dashboard HTML/CSS/JS
│
└── apps/                       # Workspace applications (one folder per app)
    └── connector-tester/
        ├── pyproject.toml
        └── src/connector_tester/
            ├── main.py
            ├── connectors/
            └── static/

Adding a New Application

1. Create a new folder under apps/:

   apps/my-new-app/
   ├── pyproject.toml
   ├── README.md
   └── src/my_new_app/
       ├── __init__.py
       └── main.py              # Must expose a run() function
   

2. Add your app's pyproject.toml with hatchling build backend and a [project.scripts] entry.

3. Register metadata in manager/src/ai_utilities_manager/app_registry.py under _APP_METADATA:

   "my-new-app": {
       "icon": "🚀",
       "color": "#f78c6c",
       "port": 8002,
       "entry_module": "my_new_app.main",
       "tags": ["example"],
   }

4. Run uv sync --all-packages and restart the manager.

Building the Executable

uv run python build.py
# Output: dist/ai-utilities (macOS ARM64 binary)

Releases

Releases are created automatically via GitHub Actions when a git tag matching v* is pushed:

git tag v0.2.0
git push origin v0.2.0

The workflow builds the macOS ARM64 binary and attaches it to the GitHub release.

Settings

User settings are stored at ~/.ai-utilities/config.json:

Setting	Default	Description
check_updates_on_startup	true	Check GitHub for updates when the manager starts
auto_update	false	Automatically download and apply updates on startup

License

MIT