Generating a test

After entering the test case you can trigger the test generation, which converts the simple language test instructions into logical steps accompanied by code to execute them using our state of the art AI models.

The final successfully generated code will be used for running the test.

Test steps are generated one-by-one, with the following logic:

• Visit the webpage is where every test case starts, opening the initial website URL configured during creation.
• Analyze current state of target application
• Logically generate next step title and description to continue with the objectives of the test instructions
• Generate (python/playwright) code for this step
• Execute generated code. If there's an error, regenerate or autonomously rework the code logic and execute.
• When the code executes without error, take a screenshot and verify if the screenshot matches the expected step outcome
• Repeat and generate steps until instructions are fulfilled, or cannot be completed (failure).