Defining Expected Results

The Expected Result is the criteria by which tapioka.ai determines if a test has passed or failed.

Evaluation Timing

The expected result is checked only after the final step of the instruction has been executed. It evaluates the final screenshot and state of the application.

Best Practices

1. Describe the Final State

The expected result should describe what the screen looks like after all actions are complete.

• Incorrect: "Data displays correctly for the Favorites, Rented, and Recordings tabs." (Since this describes intermediate steps that the AI can't verify at the end).
• Correct: "The Reminders tab is visible and contains no items."

2. Don't Over-verify

There is no need to write "The data is displayed correctly" for every step. The AI automatically checks for anomalies and errors during the entire process. Save the Expected Result for the specific outcome you want to confirm.

3. Atomic Tests

If you need to verify the content of multiple different screens, break them down into separate test cases.

Examples

• Settings View:
  • Instruction: "Go to Your Account, then Settings, and finally show the list of devices."
  • Expected Result: "A list of devices along with icons is visible on the screen."
• Content Details:
  • Instruction: "Go to the search bar, search for 'Game of Thrones', and open the series page."
  • Expected Result: "The series card is displayed. Visible elements: Add to favorites button (heart icon), rating, number of reviews, and episode list."