Halo (v1.0.0 Beta)

A translucent, liquid-glass AI copilot overlay for macOS — captures your screen, mic, and desktop content in real time to power a live AI assistant.

Note

Halo is currently in Beta (v1.0.0). Features and APIs are actively evolving.

⚡ Key Features

Liquid Glassmorphism UI — High-contrast frosted glass overlay with emerald accents that floats over your browser, IDE, or meeting apps without obscuring underlying content.
Dynamic Auto-Scaling & Resizing — Panel automatically expands to show complete AI responses and includes a manual bottom drag handle for custom window heights up to 800px.
Voice Activity Detection (VAD) — Intelligent Web Audio API energy monitoring detects spoken sentences and transcribes audio only when speech finishes, saving API quota.
Screen Analysis & Coding Assistant — Instantly captures LeetCode problems or on-screen code (Cmd+Shift+H) and delivers formatted analysis with 1-click code copying.
LaTeX Math Rendering — Beautifully formats complexity notations (e.g. $\mathcal{O}(N)$) and mathematical expressions inline and in dedicated math blocks.
Multi-Model Provider Architecture — Seamlessly stream responses from OpenAI (GPT-4o, GPT-4o-mini, Whisper) or Google Gemini (2.0 Flash, 2.0 Flash Lite, 1.5 Pro).
Global Hotkeys & Click-Through Mode — Toggle overlay visibility (Cmd+B), trigger AI assistance (Cmd+Enter), or enable mouse click-through mode for uninterrupted work.

# Clone repository
git clone https://github.com/shipitdev/halo.git
cd halo

# Install dependencies
npm install

# Start Halo
npm start

On first launch, open Settings (⚙ top right of toolbar) to enter your Gemini (AIzaSy...) or OpenAI (sk-...) API Key.

Shortcut	Action
`Cmd + B`	Toggle overlay visibility
`Cmd + Enter`	Trigger AI Assist (Screen + Transcript Context)
`Cmd + Shift + H`	Solve Code / Analyze Screen
`Cmd + Shift + X`	Quit Halo

Framework: Electron (Frameless transparent overlay)
Frontend: Vanilla HTML5, CSS3 Glassmorphism System, ES2022 JavaScript
AI Integration: OpenAI Node SDK, @google/genai Google GenAI SDK
Audio Processing: Web Audio API (RMS Analyser) + MediaRecorder WebM / PCM Audio Buffer