Heidelberg AI News

Video · GitHub ·

I made my Octolamp think with GitHub Copilot CLI hooks

In short

The video shows how GitHub Copilot CLI hooks can drive a WLED Octolamp as a physical status indicator for agent activity. It demonstrates a practical way to monitor parallel AI tasks without constantly checking terminal windows, while preserving CLI operation when the lamp is unavailable.

Key points

  • Four hooks map Copilot CLI lifecycle events to green, white, and flashing red lamp states.
  • A single HTTP POST script controls WLED colors and effects.
  • A 2-second timeout lets CLI tasks continue when the lamp is offline.
  • Removing the session-start hook resolves a race condition between white and green states.
  • User-level hooks work across CLI sessions, with project-specific customization also possible.
  • The setup can support Windows and macOS and be adapted to other WLED-compatible lights.

Written by an AI model from the video’s captions and description.

At Universe 2025, GitHub handed out Octolamps. Beatris Mendez Gandica wondered: what if hers could show when GitHub Copilot CLI is thinking? Using the native Copilot CLI hooks system, her lamp breathes green when the agent works, turns white when idle, and blinks red on errors via the WLED API.

Watch on YouTube The opening of the article at youtube.com.

In this story

A tool or model opens in the Heidelberg AI Curriculum: what it does, what it costs, how to start. A repo opens on GitHub.

Related stories

All stories