About this project

Antigravity Cockpit is a VS Code extension for monitoring quota usage of Google Antigravity AI models. The interface language follows VS Code settings and supports 16 languages. Main features: Display modes: Switch between Webview dashboard and QuickPick modes in settings. Webview provides card view, grouped/non-grouped modes, drag-and-drop sorting, and auto-grouping; QuickPick uses VS Code native APIs, suitable for environments where Webview cannot load or for users who prefer keyboard operation, supporting refresh, toggle grouping, open logs, pin models to the status bar, and rename models and groups. Status bar: Displays quota status of the currently monitored model, with six formats: icon only, status dot only, percentage only, status dot plus percentage, name plus percentage, and full display. Supports pinning multiple models; when no model is specified, automatically displays the model with the lowest remaining quota. Quota display: Each model or group shows remaining quota percentage, countdown, reset time, and progress bar. Quota sources: Supports local monitoring (reads local Antigravity client process) and authorized monitoring (accesses remote interface via authorization, independent of local process), switchable at the top right of the panel. Authorized monitoring supports multiple accounts, allowing switching the current account and viewing status. Model capability hints: Hover over model names to view capability markers such as supported input types and context window size. Grouping: Automatically or manually categorize models by quota pool, supporting custom group names, drag-and-drop sorting, and group pinning. Settings panel: Configure status bar display format, warning threshold, danger threshold, view mode (card/list), and notification toggle. User profile panel: Displays subscription level and user ID, collapsible with privacy data masking. Notifications: Send notifications when model quota falls below warning threshold or is exhausted, disableable in settings. Auto-wake: Schedule tasks to wake AI models in advance to trigger quota reset cycles, supporting daily, weekly, interval loops, and Crontab modes, with the ability to wake multiple models simultaneously. Credentials are encrypted and stored in VS Code Secret Storage, run locally, with trigger logs and AI response records provided. Usage: Click the status bar icon, press Ctrl/Cmd+Shift+Q, or run Antigravity Cockpit: Open Dashboard from the command palette to open; press Ctrl/Cmd+Shift+R to refresh when the dashboard is active. If Systems Offline appears, click Retry Connection, and click Open Logs to view debug logs. Installation: Search and install from the Open VSX marketplace, or use code --install-extension to install the VSIX file. Building from source requires Node.js v18+ and npm v9+, running npm install, npm run compile, and npm run package in sequence. The project is licensed under MIT, and the README states it is for personal learning and research use only.