Skip to main content
This guide covers common issues and their solutions when working with Seemodo.

Sandbox Issues

Sandbox not starting

1

Check provider configuration

Verify your .env file has the correct settings:
SANDBOX_PROVIDER=modal
MODAL_BACKEND_URL=https://your-backend.modal.run
2

Verify Modal deployment

Ensure your Modal backend is deployed and running:
cd modal-backend
modal deploy main.py
3

Check network connectivity

Your browser must be able to reach the Modal backend URL. Test in a new tab.
4

Review browser console

Open DevTools (F12) and check the Console tab for error messages.

Preview not loading

SymptomCauseSolution
Blank iframeVite not readyWait 5-10 seconds, then refresh
Build errorCode errorCheck AI chat for error messages
TimeoutSandbox expiredGenerate a new screen to restart
CORS errorTunnel issueTry refreshing the page

Sandbox keeps timing out

The default timeout is 60 minutes. To extend:
  1. Modify config/app.config.ts:
    modal: {
      timeoutMinutes: 120, // 2 hours
    }
    
  2. Redeploy the Modal backend

AI Generation Issues

Generation fails

Verify your AI provider keys in .env:
GEMINI_API_KEY=your_valid_key_here
Test the key by making a direct API call.
If you’re making many requests, you may hit rate limits. Wait a few minutes and try again.
Very long or complex prompts may fail. Try:
  • Breaking into smaller pieces
  • Using Plan mode for complex features
  • Being more specific about what you want
If a specific model fails, try a different one:
  • Add ?model=true to the URL
  • Select an alternative model

Poor quality output

1

Use Deep Design mode

Enable Deep Design in the sidebar for better results.
2

Add reference images

Upload screenshots or mockups to guide the AI.
3

Be more specific

Instead of “make it pretty”, specify colors, layout, and components.
4

Try a different model

Some models are better for certain types of designs.

AI not responding

If the Seemodo AI panel shows “Thinking” indefinitely:
  1. Check the connection indicator (should be green)
  2. Click the Reconnect button
  3. Refresh the page if needed
  4. Check if the sandbox is still running

Code Issues

Build errors after generation

The AI automatically detects and fixes most build errors. If an error persists:
  1. Check the error message in the AI chat
  2. Ask the AI specifically to fix it:
    Fix the TypeScript error in Button.tsx
    
  3. Review the generated code for obvious issues

Missing packages

Packages are auto-installed, but sometimes fail:
# In the sandbox terminal
pnpm add package-name
Or ask the AI:
Install lodash and use it in the utils file

HMR not working

If changes aren’t reflecting in the preview:
  1. Click “Restart Vite” in the terminal
  2. Wait for the dev server to restart
  3. Refresh the preview iframe

Seemodo Cloud Issues

Connection failed

The project reference is just the ID, not the full URL:
  • ✅ Correct: abcdefghijklmnop
  • ❌ Wrong: https://abcdefghijklmnop.supabase.co
Get a new token from Supabase Dashboard → Account → Access Tokens.Ensure it has the required permissions.
Free Supabase projects pause after inactivity. Visit the dashboard to resume.

Database operations failing

  1. Check RLS policies aren’t blocking the operation
  2. Verify the table exists in the correct schema
  3. Ask the AI to debug:
    Why is the insert into users table failing?
    

Browser Issues

Slow performance

Too many open tabs

Close other tabs, especially those with active sandboxes.

Browser cache

Clear cache and reload the page.

Extensions

Disable browser extensions that might interfere.

Hardware acceleration

Ensure hardware acceleration is enabled in browser settings.

Canvas not loading

tldraw requires WebGL support:
  1. Check that WebGL is enabled in your browser
  2. Update your graphics drivers
  3. Try a different browser (Chrome recommended)

Common Error Messages

ErrorMeaningSolution
SANDBOX_NOT_FOUNDSandbox expired or doesn’t existGenerate a new screen
SEEMODO_AI_CONNECTION_FAILEDCan’t reach Seemodo AI serverWait for sandbox to fully initialize
MODEL_UNAVAILABLEAI model is downTry a different model
RATE_LIMIT_EXCEEDEDToo many requestsWait a few minutes
INSUFFICIENT_CREDITSOut of creditsPurchase more or wait for refresh

Getting Help

If these solutions don’t resolve your issue:
  1. Check the console - Press F12 and look for error messages
  2. Export logs - Click “Export” in the Seemodo AI panel
  3. Create a minimal reproduction - Note the exact steps to reproduce
  4. Contact support - Email [email protected] with:
    • Browser and OS version
    • Steps to reproduce
    • Error messages and screenshots
    • Exported logs if available