AutoSmoke uses AI to intelligently find elements on your page, so you don't need to write CSS selectors or XPath expressions.
How It Works
When you write a step like:
Click the "Sign Up" button
Our AI:
- Analyzes the page structure
- Identifies all potential matches
- Uses context to select the correct element
- Falls back to alternative selectors if needed
Self-Healing Selectors
Traditional tests break when:
- CSS classes change
- Element IDs are updated
- DOM structure is reorganized
AutoSmoke's self-healing technology adapts to these changes automatically by:
- Using multiple selector strategies
- Prioritizing stable attributes (text, labels, roles)
- Learning from successful past executions
Providing Hints
While AutoSmoke usually finds elements correctly, you can provide hints for precision:
Using Text Content
Click the button with text "Submit Order"
Using Position
Click the first "Add to Cart" button
Click the "Delete" button next to "Product Name"
Using Section Context
In the header, click the "Login" link
In the sidebar, click "Settings"
Handling Ambiguous Elements
If multiple elements match your description:
- AutoSmoke will attempt to identify the most likely match
- If uncertain, it will ask for clarification
- You can provide more context to disambiguate
Example: Multiple Buttons
❌ Click the "Submit" button (ambiguous if multiple exist)
✅ Click the "Submit" button in the payment form
✅ Click the primary "Submit" button
Accessibility and Testing
AutoSmoke works best with accessible websites because:
- Proper labels help identify form fields
- ARIA roles indicate element purposes
- Semantic HTML provides context
This is a win-win: accessible sites are easier to test AND better for users.