Describe your test script in plain language, and the AI manual script generator powered by Sembi IQ will generate the steps for you.
Using the AI Manual Script Generator
The AI Manual Script Generator accepts a plain language description of your desired test script and suggests manual test steps automatically. Rather than writing out each step by hand, let the AI service suggest steps for you to review and accept.
Describing your test steps
The AI Manual Script Generator appears as a banner at the top of the Manual Scripts screen, with a text area. You can collapse the banner by clicking its header.
You don't need to use any special syntax. Write naturally, as you would explain the script to a colleague. Once you've finished writing your description, click Generate Steps, and the AI service will suggest a set of steps.
Example descriptions
"Test the login flow for a new user, verifying that required fields are validated and the dashboard loads after successful login."
"Verify that submitting a contact form sends a confirmation email and displays a success message. Check error handling for missing fields."
Design Guidance
If Design Guidance has been set up for your model, the AI service will use your testing goal and any uploaded context documents to improve its suggested steps. If Design Guidance is not yet configured, clicking Generate Steps will show a dialog offering two options:
Proceed without Guidance generates steps using only your description and the existing model. This can still produce useful suggestions, especially with a clearly-worded and detailed description.
Provide Guidance opens the Design Guidance setup so you can add a testing goal and context before generating.
Reviewing suggested steps
Once the AI service has processed your description, a new script tab is created, and the suggested steps appear inside this script. Each suggested step shows:
A label reading "AI Generated — pending review"
The Action, Data, and Expected Result fields, pre-filled with suggestions
Any conditions the AI service included
A banner beneath the text area shows how many AI-generated steps are pending review.
Editing suggested steps
You can click on a suggested step to enter edit mode before accepting it. In edit mode, the step expands to show the full text areas for Action, Data, and Expected Result. Parameter autocomplete is available while editing, so you can adjust parameter references as needed. Click Done to return to the preview.
Any edits you make are preserved when you accept the step.
Accepting and rejecting steps
Suggested steps can be accepted or rejected individually, or all at once.
Click the checkmark icon on a step to accept that suggestion. Upon acceptance, the step is converted into a regular script step.
Click the X icon on a step to reject that suggestion. Rejected steps are discarded.
Use Accept All to accept all remaining suggestions at once. Similarly, Reject All can discard all remaining suggestions at once, while preserving any you've already chosen to accept.
Accepted steps become regular script steps that you can further edit, reorder, and add conditions to. They are marked with an AI sparkle icon so you can identify which steps were generated entirely by the AI service, without any edits from a human. This sparkle will disappear if there are any edits to the step.
Once all suggested steps have been accepted or rejected, the script is saved automatically.
Tips for writing effective test descriptions
Be specific. The more detail you provide in your description, the more relevant the suggestions will be. Describe the user actions, the data involved, and the expected outcomes.
Use Design Guidance. If your model has a testing goal or context documents configured, the AI service can use that background to produce suggestions more relevant to your testing objectives.
Reference your parameters. If you mention specific parameter names from your model in the description, especially where you expect that a step would apply only conditionally, the AI service can more accurately map suggestions to your test data and create more accurate test steps.
If no suggestions are returned, try rephrasing your description or adding more detail. The AI service may not always be able to generate a script for the specific parameters and values in your model.