TestDriver isn’t just a computer-use agent, it’s an entire AI-native workflow for creating end-to-end tests. Follow these steps to set up your environment and get started with TestDriver.

1

Set Up TestDriver

2

Create your first test

Use the /explore command to supply test steps to the TestDriver agent. TestDriver will look at the screen and generate a test steps based on your input.

> Open Google Chrome and search for "testdriverai"

TestDriver will generate and run test steps. See the YAML Reference.

- command: focus-application
  name: Google Chrome
- command: hover-text
  text: Search Google or type a URL
  description: main search bar
  action: click
- command: type
  text: testdriverai
- command: press-keys
  keys: [enter]

Test steps are saved in the current working YAML file.

Use Control + C in the CLI or the square “stop” button in the extension to stop test execution.
3

Running tests

Once saved, you can run your test file anytime. Click the “Run” button in the extension or use the CLI command:

testdriverai run testdriver/your_test.yml

Running tests is about twice as fast as creating them, as the AI doesn’t need to interpret your input.

4

Generate self-driving tests

TestDriver can also generate tests without any input! Instead, it will suggest test steps based on what it sees on the screen.

Self-Driving Test Guide

Unleash TestDriver’s full potential by letting it create tests for you.

5

Deploy to CI/CD

The testdriverai init command automatically sets up a GitHub Action workflow for you. Any test files found in the testdriver directory will run on every push to the main branch.

GitHub Actions

Learn how to set up GitHub Actions for TestDriver.