Qubase

A complete multi-model AI assistant template for any project. Clone this template to create intelligent assistants that understand your entire codebase or knowledge base with full context.

Features

✅ Multi-project architecture with dropdown selector, assign AI Model per project
✅ Full codebase context (unlimited token budget per project)
✅ Smart file loading with automatic GitHub sync
✅ Simple UI, ChatGPT-style interface
✅ Mobile-responsive design
✅ Interactive code features (copy, define, locate, error check)
✅ Theme switching (dark/light mode)
✅ Real-time streaming responses
✅ Password protection for team access

Quick Start

1. Setup Your Project Repository

First, organize your codebase into logical folders:

your-project-repo/
├── frontend-code/          # React/Vue/etc components  
├── backend-api/            # Server code
├── documentation/          # Project docs
├── mobile-app/            # Mobile code (optional)
└── design-assets/         # UI/UX files (optional)

2. Clone This Template

git clone https://github.com/silver-ben/qubase.git my-project-assistant
cd my-project-assistant

3. Configure for Your Project

Backend Configuration

Edit backend/server-template.js:

// Update PROJECTS config (lines 47-67)
const PROJECTS = {
  'frontend': {
    name: 'Your Project Frontend',
    path: '../codebase/frontend-code',
    model: 'gpt-4.1-mini-2025-04-14',
    description: 'React frontend components and styling'
  },
  'backend': {
    name: 'Your Project Backend',
    path: '../codebase/backend-api', 
    model: 'gpt-4.1-mini-2025-04-14',
    description: 'Node.js/Python backend API code'
  },
  'docs': {
    name: 'Documentation',
    path: '../codebase/documentation',
    model: 'gpt-4.1-mini-2025-04-14', 
    description: 'Project documentation and guides'
  }
};

// Update GitHub repo config (lines 32-35)
const GITHUB_REPO = {
  owner: 'your-username',      // Your GitHub username
  repo: 'your-project-repo'    // Your project's codebase repo
};

// Update sync projects list (lines 39-43)
const SYNC_PROJECTS = [
  'frontend-code',
  'backend-api', 
  'documentation'
];

Frontend Branding

Edit frontend/src/App-template.jsx and replace these placeholders:

• {{PROJECT_NAME}} → Your Project Name
• {{PROJECT_DESCRIPTION}} → Short description
• {{FEATURE_1}} → Main feature 1
• {{FEATURE_2}} → Main feature 2
• {{FEATURE_3}} → Main feature 3
• {{FEATURE_4}} → Main feature 4
• {{DETAILED_PLACEHOLDER}} → Detailed input placeholder

Package.json Files

Update backend/package-template.json and frontend/package-template.json:

• {{PROJECT_SLUG}} → your-project-name (lowercase, dashes)
• {{PROJECT_NAME}} → Your Project Name

4. Environment Setup

# Backend setup
cd backend
cp .env-template .env
# Edit .env with your API keys
npm install

# Frontend setup  
cd ../frontend
cp package-template.json package.json
# Update placeholders in package.json
npm install

5. Development

# Terminal 1: Start backend
cd backend && node server-template.js

# Terminal 2: Start frontend  
cd frontend && npm run dev

Visit http://localhost:5173 to test your AI assistant.

6. Production Deployment

Build Frontend

cd frontend && npm run build

Deploy to Server

# Upload entire project to server
scp -r ./my-project-ai/ user@server:/var/www/ai.yourproject.com/

# Install dependencies on server
cd /var/www/ai.yourproject.com/backend && npm install
cd /var/www/ai.yourproject.com/frontend && npm install && npm run build

# Start with PM2
pm2 start server-template.js --name your-project-ai
pm2 save

Configure Web Server

Apache VirtualHost:

<VirtualHost *:443>
    ServerName ai.yourproject.com
    
    # Password Protection
    <Location />
        AuthType Basic
        AuthName "Your Project AI - Team Access"
        AuthUserFile /var/www/ai.yourproject.com/.htpasswd
        Require valid-user
    </Location>
    
    # SSL Configuration
    SSLEngine on
    SSLCertificateFile /etc/letsencrypt/live/ai.yourproject.com/fullchain.pem
    SSLCertificateKeyFile /etc/letsencrypt/live/ai.yourproject.com/privkey.pem
    
    # Reverse Proxy
    ProxyPass / http://localhost:3001/
    ProxyPassReverse / http://localhost:3001/
</VirtualHost>

Set Up Team Access

# Create password file
sudo htpasswd -c .htpasswd teamuser
# Enter secure password

# Add more users
sudo htpasswd .htpasswd developer1
sudo htpasswd .htpasswd manager1

Configuration Reference

Required Environment Variables

# OpenAI API Key (required)
OPENAI_API_KEY=sk-...

# GitHub Personal Access Token (optional, for auto-sync)
GITHUB_PAT=ghp_...

# Server Port (optional, defaults to 3001)
PORT=3001

Project Configuration

Each project in the PROJECTS object supports:

• name: Display name for the project
• path: Relative path to project files
• model: OpenAI model to use (recommend gpt-4.1-mini-2025-04-14)
• description: Brief description of the project

GitHub Integration

To enable automatic code synchronization:

1. Create a GitHub Personal Access Token with repo access
2. Add token to .env file as GITHUB_PAT
3. Update GITHUB_REPO object with your repository details
4. List folders to sync in SYNC_PROJECTS array

The system will auto-update when you push to your repository's main branch.

Cost Estimation

Per Project Setup:

• Domain: $10-15/year
• SSL: Free (Let's Encrypt)
• Server: $5-20/month (can host multiple projects)
• OpenAI API: $5-50/month (depending on usage)

Total: ~$10-85/month per project (mostly OpenAI costs)

With 500K token context per project, typical usage costs $0.15 per 1M input tokens.

File Structure

ai-assistant-template/
├── README.md                    # This file
├── backend/
│   ├── server-template.js      # Main server with placeholders
│   ├── package-template.json   # Package.json template
│   └── .env-template           # Environment variables template
├── frontend/
│   ├── src/
│   │   ├── App-template.jsx    # React app with placeholders
│   │   └── index.css           # Styling (ready to use)
│   ├── package-template.json   # Package.json template
│   └── vite.config.js          # Vite config (ready to use)
└── docs/
    └── clone-for-new-project.md # Detailed setup guide

System Prompt Customization

The AI's behavior is controlled by the system prompt in server-template.js (lines 422-437). Customize this to:

• Define your team members
• Explain your project's architecture
• Set response style and tone
• Add specific instructions for your domain

Support

For issues or questions:

1. Check the detailed guide in docs/clone-for-new-project.md
2. Test with a simple project first
3. Open an issue in this repository

Benefits

✅ Rapid deployment - 2-3 hours for new project
✅ Proven architecture - Based on production system
✅ Full codebase context - No chunking limitations
✅ Professional UX - ChatGPT-style interface
✅ Team collaboration - Shared knowledge base
✅ Auto-updating - Syncs with code changes
✅ Cost-effective - Minimal ongoing costs

Transform any codebase into an intelligent assistant that knows your entire project inside and out!

---

Created by Ben Silver | bensilver.com.au
Qubase: Query any knowledge base with full context