A GitHub Voice AI Assistant.

Jengu is an intelligent voice interface for exploring GitHub repositories. By integrating real-time voice transcription with an autonomous LLM agent (LangGraph), Jengu allows users to naturally ask for complex repository data. The AI acts as a sophisticated orchestrator—fetching data via the GitHub API, synthesizing an audio response, and simultaneously synchronizing a rich graphical overlay (charts and tables) on the user's screen.

• Voice-Driven Interaction: Speak directly to Jengu to query GitHub without touching a keyboard.
• Agentic UI Synchronization: The audio response is perfectly timed with responsive React data visualizations (Recharts).
• GitHub OAuth Integration: Securely authenticate to view private repositories and bypass public API rate limits.
• Robust Error Handling: Gracefully handles transcription ambiguities, missing repositories, and microphone permission issues.
• DigitalOcean Ready: Architected as a stateless application for seamless deployment to DigitalOcean App Platform.

• Frontend:
• Backend:
• AI Engine:
• LLM:
• Voice:
• Authentication:
• API Client:

• Node.js 18+
• A GitHub OAuth Application (Client ID & Secret)
• A GitHub Personal Access Token (PAT)
• A Vapi Assistant ID and Public Key

1. Clone the repository and install dependencies:

npm install

2. Copy the example environment variables file and fill in your keys:

cp .env.example .env

3. Run the development server:

npm run dev

Open http://localhost:3000 to interact with Jengu.

Distributed under the MIT License. See LICENSE for more information.