Skip to content

Latest commit

 

History

History
84 lines (67 loc) · 2.08 KB

File metadata and controls

84 lines (67 loc) · 2.08 KB

🚀 Dev AI OS - Setup Guide

This guide details how to run the Dev AI OS in real-time with all components connected.

Architecture Overview

  • Backend (Node.js): Port 3001. Handles API & Auth. Connected to MongoDB.
  • OS Automation (Python): Port 8000. Handles File/App operations.
  • Voice System (Python): Standalone. Processes mic input.
  • Frontend (Next.js): Port 3000. Visual Interface.

1. Prerequisites

  • Node.js (v18+)
  • Python (v3.10+)
  • MongoDB (Local or Atlas URI)
  • Google Cloud Credentials (Client ID & Secret for OAuth)

2. Integration Steps

You will need 4 separate terminal windows to run the full stack.

Terminal 1: OS Automation Service

This bridge allows the backend to control your PC.

cd apps/dev-os-automation
# Activate Python Env
..\..\env\Scripts\activate
# Start Server
python src/server.py
# ⚡ Listening on http://0.0.0.0:8000

Terminal 2: Main Backend API

The central brain linking everything.

cd apps/dev-auth-backend
# Ensure .env is configured (see .env.example)
npm run dev
# 🚀 Server listening on port 3001
# ✅ MongoDB Connected

Terminal 3: Frontend Dashboard

The visual interface.

cd apps/dev-frontend-ui
npm run dev
# 🚀 Dashboard live at http://localhost:3000

Terminal 4: Voice Listener (Optional)

If you want voice interaction.

cd apps/dev-voice-system
..\..\env\Scripts\activate
python src/main.py

3. Configuration

Backend Environment (apps/dev-auth-backend/.env)

Ensure you have a .env file with the following:

PORT=3001
MONGODB_URI=your_mongodb_connection_string
GOOGLE_CLIENT_ID=your_google_client_id
GOOGLE_CLIENT_SECRET=your_google_client_secret
GOOGLE_REDIRECT_URI=http://localhost:3000/api/auth/google/callback
JWT_ACCESS_SECRET=your_jwt_secret

4. Verification

  1. Open http://localhost:3000.
  2. Click "Initialize Core" to log in via Google.
  3. Once on the dashboard, type open notepad in the chat.
  4. Success: Notepad should open on your machine, and the command should be logged in MongoDB.