Talk to your terminal AI agent from your phone.

Tether a coding-agent CLI on your desktop to your phone over WebRTC — voice and chat, hands-free. The agent runs where your code is; you drive it from the couch, the kitchen, or the car.

Open the app

1 · Install on your desktop

macOS / Linux:

curl -fsSL https://bridle.3sln.com/install.sh | sh
Windows (PowerShell)
irm https://bridle.3sln.com/install.ps1 | iex

No API keys: speech-to-text runs on-device (Whisper), text-to-speech is your browser’s.

2 · Pair with your phone

  1. In any project, run bridle tether work claude (or … codex, etc.).
  2. It prints a QR and installs itself as a background service for that project.
  3. Scan the QR with your phone — it opens the app and links over a peer-to-peer data channel.
  4. Talk to your agent. Tap the mic to go hands-free and just speak; say “stop talking” to cut in.
Open the app Docs & source

What it’s for

  • Voice in and out — on-device Whisper STT, browser TTS, a hands-free conversation mode with barge-in.
  • Any agent — claude, codex, gemini, aider, opencode, goose… or pipe any other CLI.
  • Private and P2P — chat and files flow phone↔desktop directly; the backend only brokers the handshake.
  • Driving mode — earcons, screen wake-lock, and car/headset/lock-screen media buttons for eyes-off use.