shopify-ai-backup/QUICK_TEST.md

Quick Test Guide

🚀 Quick Deploy & Test

1. Rebuild Container (Required!)

# Stop and remove old container
docker-compose down

# Rebuild with no cache to pick up changes
docker-compose build --no-cache

# Start container
docker-compose up -d

2. Verify Container Logs Work ✅

# Tail logs - you should now see output!
docker logs -f shopify-ai-builder

# Look for lines like:
# [2024-01-11T...] Server started on http://0.0.0.0:4000
# [CONFIG] Mistral: { configured: true, ... }
# [CONFIG] Planning Settings: { ... }

Before the fix: No logs visible, all went to /var/log/chat-service.log After the fix: All logs visible in Docker logs ✅

3. Test Groq Planning ✅

Setup:

# Add to .env or docker-compose.yml
GROQ_API_KEY=your_groq_api_key_here

Test Steps:

1. Restart after adding env var: docker-compose restart
2. Go to: http://localhost:4000/admin/login
3. Navigate to "Plan models"
4. Click "Add planning model"
5. Select Provider: groq, Model: (leave empty for default)
6. Save
7. Go to builder: http://localhost:4000/apps
8. Create new project and enter a plan request
9. Check logs: docker logs shopify-ai-builder | grep "\[GROQ\]"

What to expect:

• Planning request returns a response
• Logs show: [GROQ] Starting API request
• Logs show: [GROQ] Response received { status: 200, ok: true }
• Logs show: [GROQ] Extracted reply: { replyLength: ... }

4. Test Mistral Planning ✅

Setup:

# Add to .env or docker-compose.yml
MISTRAL_API_KEY=your_mistral_api_key_here

Test Steps:

1. Restart after adding env var: docker-compose restart
2. Go to admin panel → Plan models
3. Add Mistral as planning provider
4. Go to builder and create plan request
5. Check logs: docker logs shopify-ai-builder | grep "\[MISTRAL\]"

What to expect:

• Planning request returns a response
• Logs show Mistral API request/response cycle
• Model information tracked properly

5. Verify Provider Fallback ✅

Test automatic fallback:

1. Configure multiple providers in planning chain (e.g., Groq → Mistral → OpenRouter)
2. Make a plan request
3. If first provider fails, should automatically try next
4. Check logs to see fallback chain: docker logs shopify-ai-builder | grep "\[PLAN\]"

🎯 Success Criteria

• ✅ Container logs are visible and updating in real-time
• ✅ Groq planning returns responses
• ✅ Mistral planning returns responses
• ✅ Provider fallback works automatically
• ✅ Admin panel shows all providers (openrouter, mistral, google, groq, nvidia)

🐛 Troubleshooting

Problem: Still no logs visible

Solution: Make sure you rebuilt with --no-cache

docker-compose down
docker-compose build --no-cache
docker-compose up -d

Problem: "API key not configured"

Solution: Add environment variables and restart

# Add to .env file
GROQ_API_KEY=...
MISTRAL_API_KEY=...

# Restart container
docker-compose restart

Problem: Groq returns error

Solution: Check logs for specific error message

docker logs shopify-ai-builder 2>&1 | grep -A 5 "\[GROQ\].*error"

Problem: Planning returns no response

Solution:

1. Check logs: docker logs -f shopify-ai-builder
2. Look for error messages with provider prefixes
3. Verify API key is correct
4. Try a different provider as fallback

📊 Log Examples

Good Groq Response:

[GROQ] Starting API request { url: '...', model: 'llama-3.3-70b-versatile', messageCount: 3 }
[GROQ] Response received { status: 200, ok: true }
[GROQ] Response data: { hasChoices: true, choicesLength: 1, model: '...' }
[GROQ] Extracted reply: { replyLength: 523, replyPreview: 'Here is a WordPress plugin...' }
[PLAN] Provider succeeded { provider: 'groq', model: '...', replyLength: 523 }

Good Mistral Response:

[MISTRAL] Starting API request { url: '...', model: 'mistral-large-latest', messageCount: 3 }
[MISTRAL] Response received { status: 200, ok: true }
[MISTRAL] Successfully extracted reply { replyLength: 1024, replyPreview: 'I'll help you create...' }
[PLAN] Provider succeeded { provider: 'mistral', model: 'mistral-large-latest', replyLength: 1024 }

Fallback Chain Working:

[PLAN] Trying provider { provider: 'groq', model: '' }
[GROQ] Request failed { status: 429, detail: 'Rate limit exceeded' }
[PLAN] Trying provider { provider: 'mistral', model: '' }
[MISTRAL] Successfully extracted reply { replyLength: 856 }
[PLAN] Provider succeeded { provider: 'mistral', model: 'mistral-large-latest' }

📝 Environment Variables

# Required for Groq
GROQ_API_KEY=your_groq_key_here

# Required for Mistral
MISTRAL_API_KEY=your_mistral_key_here

# Optional - use defaults if not set
GROQ_API_URL=https://api.groq.com/openai/v1/chat/completions
MISTRAL_API_URL=https://api.mistral.ai/v1/chat/completions

🔗 Quick Links

• Admin Panel: http://localhost:4000/admin/login
• Plan Models Config: http://localhost:4000/admin/plan
• Builder: http://localhost:4000/apps
• Full Documentation: See FIXES_SUMMARY.md