Run E2E tests on EAS Workflows

Set up your project

If you haven't already, create a new project and sync it with EAS.

Follow the Get started with EAS Workflows guide to create a new project and sync it with EAS. Then, configure your project and link your GitHub repository.

Add example Maestro test cases

This is what the UI of the app created from the default Expo template looks like:

Let's create two simple Maestro flows for the example app. Start by creating a directory called .maestro in the root of your project directory. This directory will contain the flows you'll configure and should be at the same level as eas.json.

Inside, create a new file called home.yml. This flow will launch the app and assert that the text "Welcome!" is visible on the home screen.

appId: dev.expo.eastestsexample # This is an example app id. Replace it with your app id.
---
- launchApp
- assertVisible: 'Welcome!'

Next, create a new flow called expand_test.yml. This flow will open the "Explore" screen in the example app, click on the "File-based routing" collapsible, and assert that the text "This app has two screens." is visible on the screen.

appId: dev.expo.eastestsexample # This is an example app id. Replace it with your app id.
---
- launchApp
- tapOn: 'Explore.*'
- tapOn: '.*File-based routing'
- assertVisible: 'This app has two screens.*'

Run Maestro tests locally (optional)

To run Maestro tests locally, install the Maestro CLI by following the instructions in Installing Maestro.

Install your app on a local Android Emulator or iOS Simulator. Open a terminal, navigate to the Maestro directory, and run the following commands to start the tests with the Maestro CLI:

The video below shows a successful run of the .maestro/expand_test.yml flow:

Build profile for E2E tests

E2E tests require a built app file: .apk for Android or .app for iOS — that EAS can install and test on an emulator/simulator.

In your eas.json file, create a build profile for E2E tests. If the file doesn't exist, run eas build:configure to generate it.

{
  "build": {
    "e2e-test": {
      "withoutCredentials": true,
      "ios": {
        "simulator": true
      },
      "android": {
        "buildType": "apk"
      }
    }
  }
}

The above build profile creates an .apk for Android and an .app for iOS. The workflow uses this profile to build the app on EAS servers.

Create an E2E test workflow

At the root of your project, create an .eas/workflows directory. Then, add a YAML file for your E2E test workflow, such as .eas/workflows/e2e-test-android.yml.

name: e2e-test-android

on:
  pull_request:
    branches: ['*'] # Run the E2E test workflow on every pull request.
jobs:
  build_android_for_e2e:
    type: build
    params:
      platform: android
      profile: e2e-test # your eas build profile for E2E test

  maestro_test:
    needs: [build_android_for_e2e]
    type: maestro
    params:
      build_id: ${{ needs.build_android_for_e2e.outputs.build_id }}
      flow_path: ['.maestro/home.yml']

This workflow builds an .apk for Android using the e2e-test build profile from the previous step. Then it runs the .maestro/home.yml flow on the built APK.

Here's an example of the same test workflow for iOS:

name: e2e-test-ios

on:
  pull_request:
    branches: ['*']

jobs:
  build_ios_for_e2e:
    type: build
    params:
      platform: ios
      profile: e2e-test # your eas build profile for E2E test

  maestro_test:
    needs: [build_ios_for_e2e]
    type: maestro
    params:
      build_id: ${{ needs.build_ios_for_e2e.outputs.build_id }}
      flow_path: ['.maestro/home.yml']

Learn more about Syntax for EAS Workflows.

Run the E2E test workflow

You can run the E2E test workflow in two ways:

1. Manually using the EAS CLI

2. Automatically when you open a pull request

The workflow uses a pull_request trigger to run automatically when someone opens a pull request to your repository. Learn more about EAS Workflow triggers.

After the workflow starts, you can track its progress and view the results in the Expo Dashboard. Here's a screenshot of a completed workflow run: