ModulesOutcomesToolsCertificateStart Here

AI-Powered E2E Testing

Write end-to-end tests for web and mobile apps that mix exact checks with AI agent goals, using the open-source e2e framework.

A short, hands-on course for developers and QA engineers who know a little TypeScript. Every screenshot comes from a real run made for the course.

5
Modules
~4
Hours
5
Hands-on labs
Web + Mobile
Targets

Course Modules

Each module has the concepts, a worked example, a lab on the same small demo app (TaskFlow), and a short knowledge check. Work through them in order.

01
Introduction and Quickstart
Why E2E tests, deterministic versus agent-driven steps, npx e2e init, the config, choosing a model, and a first green run.
~40 min
02
Core Concepts
Tests and fixtures, locators and expect, agent.act, assert and extract, how agent steps work, and the replay cache.
~40 min
03
Web Testing
Starting the app, viewports, mocking and the browser fixture, signing in with secrets, API tests, visual comparison and migrating from Playwright.
~40 min
04
Mobile Testing
iOS simulators and Android emulators with agent-device, the mobile config, device tests, permissions, real phones and hosted devices.
~40 min
05
Explore, Debug and CI
Let the agent explore for bugs, read failures and traces, keep secrets safe, run on GitHub Actions, and finish with a capstone.
~45 min

What it looks like

One mixed test: an agent goal signs up for a trial, the model judges the welcome screen, and an exact assertion checks the status. This is the first run from Module 1.

test('a visitor signs up for a trial', async ({ app, agent, screen }) => { await app.open('/'); await agent.act('sign up for a free trial as {name} with email {email}', { params: { name: 'Ada Lovelace', email: 'ada@example.test' }, }); await agent.assert('the welcome screen greets Ada by name'); await expect(screen.getByRole('status')).toContainText('trial'); });
Terminal output of a passing e2e run with AI agent steps
A real run with e2e 0.18.0 and Gemini 3.8 Flash: two tests, four model calls, 16.7 seconds.

What you will be able to do

Write resilient tests

Combine role-based locators and retrying assertions with natural-language goals, and know when each is the right tool.

Cover web and mobile

Run the same style of test in a browser, on an Android emulator or an iOS simulator, including visual comparison.

Find and fix failures

Use agent exploration to find bugs, read failure evidence and traces, and run the suite in CI with results on every pull request.

Who it is for

Developers, QA and test engineers, and students who can read basic TypeScript or JavaScript and use a terminal. No previous Playwright, Cypress or Appium experience is needed.

Tools you will use

PurposeToolNotes
RuntimeNode.js 24.8+ (or 22.22.3+)On Windows, use WSL
Test frameworke2e with @e2e-dev/web and @e2e-dev/mobileOpen source; installed per project with npm, pnpm or bun
Model for agent stepsAny AI SDK provider (we used Google Gemini)Or a ChatGPT, Copilot, OpenCode or SuperGrok subscription, or a local model
Mobile devicesAndroid SDK emulator or Xcode iOS simulatorModule 4 only; a real phone also works
CIGitHub ActionsModule 5; EAS, Bitrise and Codemagic are also supported

Certificate

Mark every module complete to unlock your course certificate.

Certificate page Start Module 1

Based on the official e2e documentation (accessed 9 October 2026). This course is independent and not affiliated with the e2e project.