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

# Create a test

Keep tests in a [test suite](/docs/platform/tests/test-suites). A suite belongs to your project and holds the tests you review and run together.

<Tabs>
  <Tab title="UI">
    1. Open **Tests** and select a suite. If you need a new suite, select **Create suite**, enter its name, and save it.
    2. In the suite, select **+ Write a test**.
    3. Enter a **Name**, a **Scenario**, and at least one **Expected behavior**.
    4. Select one or more **Personas**. Start with **Everyday caller** if it is available in your project.
    5. Add **Mock tools** or **Env** if the situation needs them, then select **Save test**.

    To edit a saved test, select its cells in the table. Text edits save when you leave the cell. Mock tools and Env open separate JSON editors with a **Save** button.
  </Tab>

  <Tab title="CLI">
    First, [install the CLI and sign in](/skills-cli-sdks/skills-and-cli#install-the-cli), then initialize your project. Pull the current tests and list the available personas:

    ```bash theme={"system"}
    egma pull
    egma persona list
    ```

    Use an existing suite, or create one:

    ```bash theme={"system"}
    egma suite create appointment-booking --name "Appointment booking"
    ```

    Create `egma/tests/appointment-booking/book-consultation.md` using the [Keep your tests in code](/docs/platform/tests/write-a-test). Use a persona name returned by `egma persona list`, then save your changes to Egma:

    ```bash theme={"system"}
    egma push
    ```

    You can also [use your coding agent to write tests](/skills-cli-sdks/skills-and-cli#install-the-skills) from your agent's prompt, tools, and code.
  </Tab>
</Tabs>

## Run your tests

A run executes the complete suite against one agent connection. Each test and selected persona combination produces a simulation. Two tests with two personas each produce four simulations.

[Start a run](/docs/platform/runs/start-and-follow-a-run), then inspect each simulation's conversation and grading results. The same suite can run against different connections, so you can check a flow through both text and voice.