A WhatsApp bridge for the Open Mind — a peer-to-peer AI that answers your friends' messages using local data, with per-person privacy.
npx @iamtrask/om-bridge
Scan the QR code with WhatsApp (Settings → Linked Devices → Link a Device). That's it.
The bridge saves incoming WhatsApp messages as JSON files in ~/Desktop/OMBox/inbox/ and polls ~/Desktop/OMBox/outbox/ for reply files to send back. Two programs talking through files.
A Python AI (running locally via Ollama) reads the inbox, generates a reply using per-person context folders, and writes it to the outbox. The bridge picks it up and sends it back through WhatsApp.
Messages must start with "om" to be processed — everything else is ignored.
If you prefer not to use npx:
git clone https://github.com/iamtrask/om-bridge.git
cd om-bridge
npm install
node bridge.js
- Blog post: Decentralized AI in 50 Lines of Python
- Video walkthrough: YouTube
- Full course: Decentralized AI from Scratch
Apache 2.0 — see LICENSE.