Archiving this project based on feedback from the Playwright team. This code is still functional/usable, and we will resurrect if there is ever a chance for deeper collaboration in the future.

Playwright CT Storybook

Test Storybook-rendered components in isolation using @playwright/test.

• Experimental package not intended for production use.
• Design / implementation borrows heavily from Playwright CT.

Introduction

Given a Storybook story (component use case):

// Button.stories.ts
import { Button } from './Button';
export default { component: Button };
export const Primary = { args: { label: 'Button', primary: true } };

CT Storybook allows you to write a Playwright test against that story using all of Playwright's features:

// Button.spec.ts
import { test, expect } from '@storybook/playwright-ct';
import * as ButtonStories from './Button.stories';

test('interacts', async ({ mount }) => {
  let count = 0;
  const button = await mount(ButtonStories.Primary, {
    onClick: () => {
      count += 1;
    },
  });
  await button.click();
  await expect(count).toBe(1);
});

Behind the scenes, this example:

1. Runs Storybook
2. Navigates to the story
3. Overrides the Button's onClick handler
4. Clicks on the button
5. Asserts the outcome

Comparison to Playwright CT

Playwright CT is Playwright's recommended way to test components in isolation. Here is the same example in Playwright CT for React:

// Button.spec.ts
import { test, expect } from '@playwright/experimental-ct-react';
import Button from './Button';

test('interacts', async ({ mount }) => {
  let count = 0;
  const button = await mount(
    <Button
      label="Button"
      primary
      onClick={() => {
        count += 1;
      }}
    />
  );
  await button.click();
  await expect(count).toBe(1);
});

There are a few key differences here:

Playwright CT React	Playwright CT Storybook
Renders component(s)	Renders story
Specifies props in test	Specifies props in story, can override in test
Renders in CT dev server	Renders in Storybook dev server
Compatible with Vite	Compatible with Vite, Webpack, Rspack, etc.
Compatible with React, Solid, Svelte, Vue	Compatible with CT renderers + Angular, Web components, Qwik
Limitations on props, imports	No known limitations

Playwright CT lacks the concept of stories, so the tests render the components directly. This has the benefit of being able to see the component setup and test all in one place, but leads to some limitations due to the fact that the test file is mixing Node and Browser code in a single file.

You should consider using Playwright CT Storybook over Playwright CT if:

• You already use Storybook for component development and documentation
• You're using a renderer that is not supported by Playwright CT (e.g. Angular)
• You're hitting up against any of the limitations of CT (e.g. test imports)

Comparison to Storybook Test Runner

Storybook also provides component testing with its Test Runner (that also uses Playwright under the hood).

Here's the same example using Storybook's play function:

// Button.stories.ts
import { within, userEvent } from '@storybook/testing-library';
import { expect } from '@storybook/jest';

import { Button } from './Button';
export default { component: Button }

let count = 0;
export const Interaction {
  args: { label: 'Button', primary: true, onClick: () => { count += 1 } },
  play: async ({ canvasElement }) => {
    await userEvent.click(within(canvasElement).getByRole('button'));
    await expect(count).toBe(1);
  }
}

Storybook Test Runner	Playwright CT Storybook
Tests run in the browser	Tests run in node
Runs play function	Runs play function if present AND test function
Share & debug test in the browser	Run and debug test in Playwright UI / debugger
No need to write extra tests	Write a test alongside your story as needed
Interact using Testing Library	Interact using all of Playwright facilities
Compatible with Chromatic & SB ecosystem	Incompatible

The main difference between Storybook Test Runner and Playwright CT is that Storybook's play functions are entirely browser-based. This means that you can publish and inspect your tests in a static build of Storybook in the cloud, which is great for collaboration debugging CI failures. The drawback of this approach is that some of Playwright's features, such as the ability to wait for network idle, are not unavailble in "user-land" browser code.

You should use Playwright CT Storybook over Storybook Test Runner if:

• You prefer Playwright's test syntax
• You prefer Playwright's workflow (UI, debugger, etc)
• You need Playwright's richer facilities for interacting with the browser such as:
  • wait for network idle
  • truly force element pseudostates
  • evaluate arbitrary test code in the browser

Installation and setup

Assuming you have a project that already uses Storybook.

First install Playwright:

npm init playwright@latest

Then install this package:

npm install @storybook/playwright-ct

Then update your Playwright config to use @storybook/playwright-ct's defineConfig:

- import { defineConfig, devices } from '@playwright/test';
+ import { defineConfig, devices } from '@storybook/playwright-ct';

This is a thin wrapper around Playwright's defineConfig that tells Playwright to run against your Storybook dev server.

Then run Playwright to verify the installation so far:

npm run playwright test

This will run the example E2E tests that Playwright installs in your project.

Then, you can write your first test based on your existing stories. This test uses Button.stories that is included as an example file in the Storybook installation, but it can be any Story file.

import { test, expect } from '@storybook/playwright-ct';
import * as ButtonStories from './Button.stories';

test('renders', async ({ mount }) => {
  const button = await mount(ButtonStories.Primary);
  await expect(button).toContainText('Button');
});

To execute the test run:

npm run playwright test

This will execute the test you have just written alongside the E2E tests from before.

TODO

• beforeMount / afterMount hooks unimplemented
• How to use this alongside play function
• How to use this alongside Playwright E2E