Test
Outcome: After this step, you've verified end-to-end that inputs reach your webhook and results return via callback in the correct format.
The Test step lets you run a complete end-to-end test of your AI Short before submitting for review. This ensures your automation works correctly and delivers results in the expected format.
Overview
Testing creates a "test order" that simulates the complete buyer experience – from filling out inputs to receiving results. This helps you catch issues before your AI Short goes live.
How testing works
- Fill out inputs – Use the same form buyers will see
- Submit test order – We send inputs to your webhook
- Processing starts – Your automation receives and processes the data
- Status polling – We check every 2 seconds for completion
- Results display – Output appears when your automation sends callback
Test environment features
Real input forms
- Identical to production – Same form buyers will use
- All input types supported – Text, files, dates, selections, etc.
- File uploads – Test with real files if your AI Short processes them
- Validation – Form validation works exactly like for buyers
Live processing
- Real webhook calls – We send actual data to your automation
- Production headers – Same authentication headers as live orders
- Real callback URL – Your automation sends results to our callback endpoint
- Status tracking – See processing status updates in real-time
Supported result types
Text and data
- text – Plain text displays with formatting preserved
- html – Rendered with styling in a preview window
- json – Formatted and syntax-highlighted for readability
- csv – Displayed as a table with sortable columns
Files and documents
- file – Download link with file type detection
- xlsx – Excel files with download and preview options
- url – Clickable links to external resources
Multi-file results
- File gallery – Visual grid of multiple files
- Individual downloads – Download each file separately
- ZIP downloads – "Download All as ZIP" for convenience
- File details – Names, types, and sizes displayed
Testing best practices
Use realistic inputs
- Real data – Test with actual websites, product names, or data files
- Edge cases – Try unusual inputs to see how your automation handles them
- File variety – Upload different file types and sizes if applicable
- Required fields – Verify required field validation works
Verify complete workflow
- Input processing – Confirm your automation receives all input data
- Callback delivery – Ensure results are sent back successfully
- Output format – Check that results match your expected format
- Error handling – Test what happens when processing fails
Common test scenarios
SEO Analysis Agent test
- Input:
website_url: https://example.com - Expected: HTML report with SEO recommendations
- Check: Report contains analysis sections and actionable advice
Product Research Agent test
- Input:
search_term: wireless headphones, platform_context: "Amazon"(context only; results are used manually) - Expected: CSV file with product data
- Check: CSV downloads correctly and contains price/rating data
Content Audit Agent test
- Input: Upload PDF file with marketing content
- Expected: Text report with grammar and style feedback
- Check: Report analyzes uploaded content specifically
Troubleshooting test issues
Processing stuck or failed
- Check webhook logs – Verify your automation received the request
- Validate callback – Ensure you're sending results to the correct URL
- Check callback secret – Verify you're including the
x-callback-secretheader - Review error logs – Look for processing errors in your automation
Results not displaying correctly
- Output format – Confirm you're sending the correct output type
- File encoding – Use proper base64 encoding for file results
- JSON structure – Validate callback payload matches expected format
- Content type – Ensure proper MIME types for file results
Important notes
💡 Test environment isolation
- • Test orders don't affect your earnings or analytics
- • Test data isn't stored permanently in our system
- • Test orders don't appear in customer order lists
- • Use test orders freely to refine your AI Short
Callback requirements
- Required header – Must include
x-callback-secret - Order ID – Include the order ID we sent in your webhook payload
- Success response – Use
ok: truefor successful results - Error handling – Use
ok: falsewith error details for failures
Tips for successful testing
- Test multiple times – Run several tests with different inputs
- Monitor your logs – Check your automation platform's execution logs
- Test file handling – If your AI Short processes files, upload various types
- Verify timing – Ensure processing completes within your estimated ETA
- Document issues – Keep notes on any problems you discover
What's next?
Once testing confirms your AI Short works correctly, you'll proceed to the final Review & Submit step where you can preview your listing and submit for admin approval.
[SCREENSHOT HERE: Test interface showing input form and results]