> ## Documentation Index
> Fetch the complete documentation index at: https://documents.tmsone.app/llms.txt
> Use this file to discover all available pages before exploring further.

# AI Test Generation

TMS ONE uses AI to automatically generate test cases from your acceptance criteria. This is also powerful for **Product Owners and Business Analysts** who want to turn user stories and requirements into ready-to-use test cases — without needing to write them manually.

***

## How it works

You provide a description of the feature or user story, and the AI generates:

* A **Gherkin style scenario** (Given / When / Then format) for human readability
* **Executable step definition code** in your chosen programming language (JavaScript, Python, Java, or C#)

Both outputs can be reviewed, edited, and saved directly into your project.

***

## Step-by-step guide

### 1. Navigate to AI Test Generation

From the left sidebar, click **AI Test Generation** under your project.

### 2. Fill in the generation form

<ParamField path="Test Case Title" type="text" required>
  A name for the test case you want to generate.\
  Example: *"User login with valid credentials"*
</ParamField>

<ParamField path="Test Description & Acceptance Criteria" type="text" required>
  Describe the feature or scenario in plain language. The more detail you provide, the better the AI output will be.

  Example:

  ```text theme={null}
  Users should be able to log in using a valid email and password.
  On successful login they should be redirected to the dashboard.
  On failure, an error message should be displayed.
  ```
</ParamField>

<ParamField path="Module" type="select" required>
  Select which module this test case belongs to. This gives the AI context about the area of the application being tested.
</ParamField>

<ParamField path="Priority" type="select" required>
  Set the priority level for the generated test case.
</ParamField>

<ParamField path="Tags" type="multiselect">
  Optional. Add one or more tags to categorise the test case (e.g. `regression`, `smoke`, `ui`).
</ParamField>

### 3. Generate the test case

Click **Generate Test Case**. The AI will process your input and return results in the right panel.

<Tip>
  If the first result isn't quite right, click **Regenerate** to try again with the same inputs.
</Tip>

***

## Understanding the output

Once generated, you'll see two tabs on the right:

### Gherkins Preview tab

Displays the test case in **Gherkin format** — a plain English, structured format using Given / When / Then syntax.

```gherkin theme={null}
Feature: User Login

  Scenario: Successful login with valid credentials
    Given the user is on the login page
    When the user enters a valid email and password
    And clicks the Login button
    Then the user should be redirected to the dashboard
```

This format is readable by both technical and non-technical stakeholders, and can be used directly in BDD frameworks like Cucumber or SpecFlow.

### Test Code tab

Displays executable step definition code in your chosen language. You can switch languages at any time — the code will regenerate automatically while keeping the Gherkin preview unchanged.

***

## Editing the output

Before saving, you can refine the generated test case:

**To edit the Gherkin preview:**

1. Click the **Edit** button in the top right of the results panel
2. Modify the text directly in the editor
3. Click **Save Edit** to apply your changes

**To change the language:**

* Use the language selector in the **Test Code** tab — the code regenerates automatically

**To start over:**

* Click **Reset** on the left form panel to clear everything and start fresh

<Warning>
  If you navigate away without saving, your generated test case will be lost. Make sure to save before leaving the page.
</Warning>

***

## Saving to your project

When you're happy with the output, click **Save Test Case**. The test case will be saved to the selected module in your project and will appear in your test cases list.

You can also:

* **Copy** the current tab content to clipboard
* **Download** the test code as a file (`.js`, `.py`, `.java`, or `.cs`)

***

## Tips for better AI output

<AccordionGroup>
  <Accordion title="Write detailed acceptance criteria">
    Vague inputs produce vague outputs. Include specific details about what the user does, what the system should do, and what the expected result is. Think of it as a user story with clear acceptance criteria.
  </Accordion>

  <Accordion title="Include positive and negative scenarios">
    Mention both happy path and failure scenarios in your description. For example: *"valid credentials should redirect to dashboard; invalid credentials should show an error message"*. The AI will generate separate scenarios for each.
  </Accordion>

  <Accordion title="Select the right module">
    The module provides context to the AI. Selecting the correct module (e.g. Login, Checkout, Dashboard) helps the AI generate more relevant test cases.
  </Accordion>

  <Accordion title="Regenerate if needed">
    AI outputs can vary. If the first result isn't quite right, click **Regenerate** to get a fresh attempt with the same inputs. You can also tweak your acceptance criteria and regenerate.
  </Accordion>
</AccordionGroup>

***

## Permissions

Access to AI test generation is role-based:

| Action                   | Who can do it                         |
| ------------------------ | ------------------------------------- |
| Generate test cases      | Users with Generate permission        |
| Regenerate               | Users with Regenerate permission      |
| Edit the Gherkin preview | Admins and users with Edit permission |
| Save to project          | Users with Save permission            |

Contact your Admin if you don't have access to this feature.

***

## What's next?

<CardGroup cols={2}>
  <Card title="Running Tests" icon="play" href="/qa/running-tests">
    Execute your generated test cases in a test run
  </Card>

  <Card title="Writing Test Cases Manually" icon="pen" href="/qa/writing-test-cases">
    Prefer to write test cases yourself? Here's how
  </Card>
</CardGroup>