DEPLOYMENT.md

# 🚀 Deployment Guide: מראות to Hugging Face Spaces

This guide will help you deploy the מראות (Mirrors) application to your Hugging Face Space.

## 📋 Pre-deployment Checklist

### Files Ready for Upload:
- ✅ `app.py` - Main application (optimized for HF Spaces)
- ✅ `requirements.txt` - Dependencies
- ✅ `README.md` - Documentation  
- ✅ `prompt_engineering.py` - Hebrew personas module
- ✅ `conversation_manager.py` - Session management
- ✅ `app_config.py` - Configuration for HF deployment
- ✅ `.gitignore` - Ignore unnecessary files

## 🌐 Deployment Steps

### Option 1: Git Upload (Recommended)

1. **Clone your HF Space repository:**
   ```bash
   git clone https://huggingface.co/spaces/shim5/mirrors
   cd mirrors
   ```

2. **Copy application files:**
   ```bash
   # Copy all the Python files
   cp /path/to/your/project/*.py .
   cp /path/to/your/project/requirements.txt .
   cp /path/to/your/project/README.md .
   cp /path/to/your/project/.gitignore .
   ```

3. **Commit and push:**
   ```bash
   git add .
   git commit -m "Deploy מראות Hebrew Self-Reflective AI Agent"
   git push
   ```

### Option 2: Direct Web Upload

1. Go to https://huggingface.co/spaces/shim5/mirrors
2. Click "Files" tab
3. Upload each file individually:
   - `app.py`
   - `requirements.txt`
   - `README.md`
   - `prompt_engineering.py`
   - `conversation_manager.py`
   - `app_config.py`

## 🔧 HF Spaces Configuration

Your Space should be configured as:
- **SDK**: Gradio
- **Python Version**: 3.9+
- **Hardware**: CPU Basic (free tier)
- **Visibility**: Public

## 🎯 Expected Behavior

1. **Build Time**: ~3-5 minutes for dependency installation
2. **First Load**: May take 1-2 minutes to load the Hebrew model
3. **Fallback Mode**: If model loading fails, app will run in demo mode
4. **Performance**: Response time ~2-5 seconds per message

## 🐛 Troubleshooting

### Common Issues:

1. **Build Fails**:
   - Check `requirements.txt` formatting
   - Ensure all Python files are valid syntax

2. **Model Loading Issues**:
   - App will automatically fall back to demo mode
   - Check logs for specific errors

3. **Hebrew Text Issues**:
   - Ensure browser supports RTL text
   - Check CSS is loading correctly

### Debug Mode:

To enable debug logging, modify `app_config.py`:
```python
HF_SPACES_CONFIG = {
    "debug": True,
    # ... other settings
}
```

## 📊 Monitoring

After deployment:
1. Check the build logs in HF Spaces
2. Test the 3-step user flow
3. Verify Hebrew text displays correctly
4. Test conversation with different parts

## 🎉 Success Indicators

✅ App loads without errors  
✅ Hebrew interface displays correctly  
✅ All 5 psychological parts are selectable  
✅ Conversations flow smoothly  
✅ User can customize persona names  
✅ Session state persists during conversation  

## 🔄 Updates

To update the deployed app:
1. Modify files locally
2. Re-upload to HF Space
3. Space will automatically rebuild

---

**Ready to deploy! 🚀**