Run end-to-end tests

Run browser tests against Browserless without installing or managing Chrome locally.

Prerequisites

A Browserless API token from your account dashboard

Steps

AI Agent
REST API
Frameworks
BQL

Use the Browserless MCP server to run end-to-end tests from any MCP-compatible AI agent (Claude Desktop, Cursor, Windsurf, ChatGPT, etc.).

1. Connect the MCP server

Send this prompt to your AI agent to install the Browserless MCP server:

Go to https://github.com/browserless/browserless-mcp/blob/main/install.md
and follow the instructions to install the Browserless MCP server
for my client.

2. Run an E2E test

Use browserless_agent. E2E testing requires multi-step interaction with the page.

Use the browserless_agent tool to go to https://scraping-sandbox.netlify.app/products,
verify the page loads correctly, check the product catalog heading
is visible, click on a product, and confirm the product details appear

The /function endpoint lets you send a browser script as the request body and get structured JSON back. It's the fastest way to drop a smoke test into a CI pipeline without adding a framework dependency. You're just sending HTTP and checking a passed boolean.

cURL
JavaScript
Python
Java
C#

View Full Code on GitHub

1. Write the test script

The /function endpoint accepts a JavaScript module as the request body. The exported function receives a page object and can return any JSON-serializable value:

2. Send the request

curl -X POST \
  "https://production-sfo.browserless.io/function?token=YOUR_API_TOKEN_HERE" \
  -H "Content-Type: application/javascript" \
  --data-raw 'module.exports = async ({ page }) => {
    await page.goto("https://scraping-sandbox.netlify.app/products");

    const title = await page.title();
    const heading = await page.$eval("h1", el => el.textContent.trim());
    const hasLink = (await page.$("a[href]")) !== null;

    const results = [
      { test: "title includes Product Catalog", passed: title.includes("Product Catalog") },
      { test: "h1 reads Product Catalog", passed: heading === "Product Catalog" },
      { test: "page has at least one link", passed: hasLink },
    ];

    return { passed: results.every(r => r.passed), results };
  };'

3. Check the output

{
  "passed": true,
  "results": [
    { "test": "title includes Product Catalog", "passed": true },
    { "test": "h1 reads Product Catalog", "passed": true },
    { "test": "page has at least one link", "passed": true }
  ]
}

View Full Code on GitHub

1. Send the request

const code = `module.exports = async ({ page }) => {
  await page.goto('https://scraping-sandbox.netlify.app/products');

  const title = await page.title();
  const heading = await page.$eval('h1', el => el.textContent.trim());
  const hasLink = (await page.$('a[href]')) !== null;

  const results = [
    { test: 'title includes Product Catalog', passed: title.includes('Product Catalog') },
    { test: 'h1 reads Product Catalog', passed: heading === 'Product Catalog' },
    { test: 'page has at least one link', passed: hasLink },
  ];

  return { passed: results.every(r => r.passed), results };
};`;

const response = await fetch(
  'https://production-sfo.browserless.io/function?token=YOUR_API_TOKEN_HERE',
  {
    method: 'POST',
    headers: { 'Content-Type': 'application/javascript' },
    body: code,
  }
);

const { passed, results } = await response.json();
results.forEach(r => console.log(`${r.passed ? '✓' : '✗'} ${r.test}`));
if (!passed) process.exit(1);

2. Check the output

Run with node e2e-testing.mjs. Each test prints a pass or fail line. The process exits with code 1 if any test fails, so CI systems pick up the failure automatically.

View Full Code on GitHub

1. Install dependencies

pip install requests

2. Send the request

import sys
import requests

code = """
module.exports = async ({ page }) => {
  await page.goto('https://scraping-sandbox.netlify.app/products');

  const title = await page.title();
  const heading = await page.$eval('h1', el => el.textContent.trim());
  const hasLink = (await page.$('a[href]')) !== null;

  const results = [
    { test: 'title includes Product Catalog', passed: title.includes('Product Catalog') },
    { test: 'h1 reads Product Catalog', passed: heading === 'Product Catalog' },
    { test: 'page has at least one link', passed: hasLink },
  ];

  return { passed: results.every(r => r.passed), results };
};
"""

response = requests.post(
    'https://production-sfo.browserless.io/function?token=YOUR_API_TOKEN_HERE',
    headers={'Content-Type': 'application/javascript'},
    data=code.encode('utf-8'),
)

data = response.json()
for result in data['results']:
    icon = '✓' if result['passed'] else '✗'
    print(f"{icon} {result['test']}")

if not data['passed']:
    sys.exit(1)

3. Check the output

Run with python e2e_testing.py. The process exits with code 1 if any test fails.

View Full Code on GitHub

1. Send the request

import java.net.URI;
import java.net.http.*;

String token = "YOUR_API_TOKEN_HERE";
String endpoint = "https://production-sfo.browserless.io/function?token=" + token;

String code = """
    module.exports = async ({ page }) => {
      await page.goto('https://scraping-sandbox.netlify.app/products');
      const title = await page.title();
      const heading = await page.$eval('h1', el => el.textContent.trim());
      const hasLink = (await page.$('a[href]')) !== null;
      const results = [
        { test: 'title includes Product Catalog', passed: title.includes('Product Catalog') },
        { test: 'h1 reads Product Catalog', passed: heading === 'Product Catalog' },
        { test: 'page has at least one link', passed: hasLink },
      ];
      return { passed: results.every(r => r.passed), results };
    };
    """;

HttpClient client = HttpClient.newHttpClient();
HttpRequest request = HttpRequest.newBuilder()
    .uri(URI.create(endpoint))
    .header("Content-Type", "application/javascript")
    .POST(HttpRequest.BodyPublishers.ofString(code))
    .build();

HttpResponse<String> response = client.send(request, HttpResponse.BodyHandlers.ofString());
System.out.println(response.body());

2. Check the output

Run the class. The response body is JSON with a passed boolean and a results array containing one entry per assertion.

View Full Code on GitHub

1. Send the request

using System.Net.Http;
using System.Text;

string token = "YOUR_API_TOKEN_HERE";
string endpoint = $"https://production-sfo.browserless.io/function?token={token}";

string code = @"module.exports = async ({ page }) => {
  await page.goto('https://scraping-sandbox.netlify.app/products');
  const title = await page.title();
  const heading = await page.$eval('h1', el => el.textContent.trim());
  const hasLink = (await page.$('a[href]')) !== null;
  const results = [
    { test: 'title includes Product Catalog', passed: title.includes('Product Catalog') },
    { test: 'h1 reads Product Catalog', passed: heading === 'Product Catalog' },
    { test: 'page has at least one link', passed: hasLink },
  ];
  return { passed: results.every(r => r.passed), results };
};";

using (HttpClient httpClient = new HttpClient())
{
    var content = new StringContent(code, Encoding.UTF8, "application/javascript");
    var response = await httpClient.PostAsync(endpoint, content);
    string result = await response.Content.ReadAsStringAsync();
    Console.WriteLine(result);
}

2. Check the output

Run the program. The response body is JSON with a passed boolean and a results array containing one entry per assertion.

Use a browser framework when you need a full assertion library, multi-step flows, or parallel test execution. Frameworks give you retrying assertions, test lifecycle hooks, and reporters that the raw REST API doesn't provide.

Puppeteer
Playwright (JS)
Playwright (Python)
Playwright (Java)
Playwright (C#)

View Full Code on GitHub

1. Install dependencies

npm install puppeteer-core

2. Connect and run assertions

import puppeteer from 'puppeteer-core';

const browser = await puppeteer.connect({
  browserWSEndpoint: 'wss://production-sfo.browserless.io?token=YOUR_API_TOKEN_HERE',
});

const results = [];

try {
  const page = await browser.newPage();
  await page.goto('https://scraping-sandbox.netlify.app/products', { waitUntil: 'networkidle2' });

  const title = await page.title();
  results.push({
    test: 'title includes "Product Catalog"',
    passed: title.includes('Product Catalog'),
  });

  const heading = await page.$eval('h1', el => el.textContent.trim());
  results.push({
    test: 'h1 reads "Product Catalog"',
    passed: heading === 'Product Catalog',
  });

  const link = await page.$('a[href]');
  results.push({
    test: 'page has at least one link',
    passed: link !== null,
  });
} finally {
  await browser.close();
}

const failed = results.filter(r => !r.passed);
results.forEach(r => console.log(`${r.passed ? '✓' : '✗'} ${r.test}`));
if (failed.length) {
  console.error(`\n${failed.length} test(s) failed`);
  process.exit(1);
}

3. Check the output

Run with node e2e-testing.mjs. The process exits with code 1 if any assertion fails, so your CI system registers it as a failure.

✓ title includes "Product Catalog"
✓ h1 reads "Product Catalog"
✓ page has at least one link

View Full Code on GitHub

1. Install dependencies

npm install @playwright/test playwright-core

2. Configure Browserless

Add a playwright.config.js file. The wsEndpoint uses /chromium/playwright rather than a plain CDP endpoint because Playwright's connection protocol requires the dedicated Playwright path. Plain CDP won't negotiate the right capabilities. Set workers to control parallelism, retries: 1 so flaky tests get one retry, and trace: 'on-first-retry' so Playwright captures a trace zip only on that retry. Skipping it on the first pass keeps things fast, and you only need it when something actually fails:

import { defineConfig } from '@playwright/test';

export default defineConfig({
  workers: 4,
  retries: 1,
  use: {
    trace: 'on-first-retry',
    connectOptions: {
      wsEndpoint: 'wss://production-sfo.browserless.io/chromium/playwright?token=YOUR_API_TOKEN_HERE',
    },
  },
});

3. Write tests

Playwright's expect() assertions are web-first. They poll the DOM until the condition passes or the timeout expires, so you don't need manual waits. Use page.route() inside individual tests to intercept or mock specific network requests without affecting other tests:

// tests/example.spec.js
import { test, expect } from '@playwright/test';

test('page has correct title', async ({ page }) => {
  await page.goto('https://scraping-sandbox.netlify.app/products');
  await expect(page).toHaveTitle(/Product Catalog/);
});

test('heading is visible and correct', async ({ page }) => {
  await page.goto('https://scraping-sandbox.netlify.app/products');
  const heading = page.locator('h1');
  await expect(heading).toBeVisible();
  await expect(heading).toHaveText('Product Catalog');
});

test('page contains a link', async ({ page }) => {
  await page.goto('https://scraping-sandbox.netlify.app/products');
  await expect(page.locator('a[href]').first()).toBeVisible();
});

test('page loads without analytics calls', async ({ page }) => {
  const blocked = [];
  await page.route('**/analytics/**', route => {
    blocked.push(route.request().url());
    route.abort();
  });
  await page.goto('https://scraping-sandbox.netlify.app/products');
  await expect(page).toHaveTitle(/Product Catalog/);
  console.log(`Blocked ${blocked.length} analytics request(s)`);
});

4. Run the tests

npx playwright test

5. Cross-browser testing

Browserless exposes separate endpoints for Chromium, Firefox, and WebKit. Use Playwright's projects array to target all three from the same config. Each project connects to its own endpoint and Playwright runs them as distinct test suites:

import { defineConfig } from '@playwright/test';

const token = 'YOUR_API_TOKEN_HERE';

export default defineConfig({
  workers: 4,
  retries: 1,
  use: { trace: 'on-first-retry' },
  projects: [
    {
      name: 'chromium',
      use: {
        connectOptions: {
          wsEndpoint: `wss://production-sfo.browserless.io/chromium/playwright?token=${token}`,
        },
      },
    },
    {
      name: 'firefox',
      use: {
        connectOptions: {
          wsEndpoint: `wss://production-sfo.browserless.io/firefox/playwright?token=${token}`,
        },
      },
    },
    {
      name: 'webkit',
      use: {
        connectOptions: {
          wsEndpoint: `wss://production-sfo.browserless.io/webkit/playwright?token=${token}`,
        },
      },
    },
  ],
});

Run with npx playwright test. Playwright sends each test to the matching Browserless browser.

6. Debug failed tests with Trace Viewer

With trace: 'on-first-retry' set, Playwright writes a trace.zip to test-results/ whenever a test fails its first run. Open it locally:

npx playwright show-trace test-results/<test-name>/trace.zip

The Trace Viewer gives you a step-by-step timeline with DOM snapshots, network requests, and console logs. No SSH or access to the remote machine needed.

7. Check the output

Running 4 tests using 4 workers

  ✓  [chromium] › example.spec.js:3 › page has correct title (1.2s)
  ✓  [chromium] › example.spec.js:8 › heading is visible and correct (0.9s)
  ✓  [chromium] › example.spec.js:14 › page contains a link (0.7s)
  ✓  [chromium] › example.spec.js:20 › page loads without analytics calls (1.1s)

  4 passed (1.5s)

View Full Code on GitHub

1. Install dependencies

pip install pytest pytest-playwright playwright pytest-xdist
playwright install

2. Configure Browserless

Override the browser fixture in conftest.py to connect to Browserless instead of launching a local browser. The scope="session" means a single browser connection is shared across all tests in the run. Individual tests each get their own page from the default page fixture:

# conftest.py
import pytest
from playwright.sync_api import Playwright

@pytest.fixture(scope="session")
def browser(playwright: Playwright):
    browser = playwright.chromium.connect(
        "wss://production-sfo.browserless.io/chromium/playwright?token=YOUR_API_TOKEN_HERE"
    )
    yield browser
    browser.close()

3. Write tests

# tests/test_example.py
import re
from playwright.sync_api import expect

def test_page_has_correct_title(page):
    page.goto("https://scraping-sandbox.netlify.app/products")
    expect(page).to_have_title(re.compile(r"Product Catalog"))

def test_heading_is_visible(page):
    page.goto("https://scraping-sandbox.netlify.app/products")
    heading = page.locator("h1")
    expect(heading).to_be_visible()
    expect(heading).to_have_text("Product Catalog")

def test_page_has_link(page):
    page.goto("https://scraping-sandbox.netlify.app/products")
    expect(page.locator("a[href]").first).to_be_visible()

def test_page_loads_without_analytics(page):
    blocked = []

    def handle_route(route):
        blocked.append(route.request.url)
        route.abort()

    page.route("**/analytics/**", handle_route)
    page.goto("https://scraping-sandbox.netlify.app/products")
    expect(page).to_have_title(re.compile(r"Product Catalog"))
    print(f"Blocked {len(blocked)} analytics request(s)")

4. Run the tests

pytest tests/ -n 4

The -n 4 flag (from pytest-xdist) runs 4 tests in parallel, each getting its own Browserless session.

5. Cross-browser testing

Parameterize the browser fixture so every test runs against Chromium, Firefox, and WebKit. Each parameterized value produces a separate browser instance connecting to its own endpoint:

# conftest.py
import pytest
from playwright.sync_api import Playwright

TOKEN = "YOUR_API_TOKEN_HERE"
ENDPOINTS = {
    "chromium": f"wss://production-sfo.browserless.io/chromium/playwright?token={TOKEN}",
    "firefox":  f"wss://production-sfo.browserless.io/firefox/playwright?token={TOKEN}",
    "webkit":   f"wss://production-sfo.browserless.io/webkit/playwright?token={TOKEN}",
}

@pytest.fixture(scope="session", params=["chromium", "firefox", "webkit"])
def browser(playwright: Playwright, request):
    name = request.param
    b = getattr(playwright, name).connect(ENDPOINTS[name])
    yield b
    b.close()

6. Debug with Trace Viewer

Pass --tracing retain-on-failure to capture traces only for failing tests, so you're not storing trace zips for every passing run:

pytest --tracing retain-on-failure

Traces are saved to test-results/. Open one with:

playwright show-trace test-results/<test-name>/trace.zip

7. Check the output

PASSED tests/test_example.py::test_page_has_correct_title
PASSED tests/test_example.py::test_heading_is_visible
PASSED tests/test_example.py::test_page_has_link
PASSED tests/test_example.py::test_page_loads_without_analytics

4 passed in 2.3s

View Full Code on GitHub

1. Install dependencies

Add to pom.xml:

<dependency>
    <groupId>com.microsoft.playwright</groupId>
    <artifactId>playwright</artifactId>
    <version>1.40.0</version>
    <scope>test</scope>
</dependency>
<dependency>
    <groupId>org.junit.jupiter</groupId>
    <artifactId>junit-jupiter</artifactId>
    <version>5.10.0</version>
    <scope>test</scope>
</dependency>

2. Set up the test class

Connect to Browserless once in @BeforeAll so the WebSocket handshake only happens once per suite. Open a fresh context per test in @BeforeEach. Separate contexts give each test isolated cookies, storage, and network state, so tests can't bleed state into each other:

import com.microsoft.playwright.*;
import org.junit.jupiter.api.*;

class ExampleTest {
    static Playwright playwright;
    static Browser browser;
    BrowserContext context;
    Page page;

    @BeforeAll
    static void connect() {
        playwright = Playwright.create();
        browser = playwright.chromium().connect(
            "wss://production-sfo.browserless.io/chromium/playwright?token=YOUR_API_TOKEN_HERE"
        );
    }

    @AfterAll
    static void disconnect() {
        playwright.close();
    }

    @BeforeEach
    void createContextAndPage() {
        context = browser.newContext();
        page = context.newPage();
    }

    @AfterEach
    void closeContext() {
        context.close();
    }
}

3. Write tests

import static com.microsoft.playwright.assertions.PlaywrightAssertions.assertThat;
import java.util.*;
import java.util.regex.Pattern;

@Test
void pageHasCorrectTitle() {
    page.navigate("https://scraping-sandbox.netlify.app/products");
    assertThat(page).hasTitle(Pattern.compile("Product Catalog"));
}

@Test
void headingIsVisible() {
    page.navigate("https://scraping-sandbox.netlify.app/products");
    Locator heading = page.locator("h1");
    assertThat(heading).isVisible();
    assertThat(heading).hasText("Product Catalog");
}

@Test
void pageHasLink() {
    page.navigate("https://scraping-sandbox.netlify.app/products");
    assertThat(page.locator("a[href]").first()).isVisible();
}

@Test
void pageLoadsWithoutAnalytics() {
    List<String> blocked = new ArrayList<>();
    page.route("**/analytics/**", route -> {
        blocked.add(route.request().url());
        route.abort();
    });
    page.navigate("https://scraping-sandbox.netlify.app/products");
    assertThat(page).hasTitle(Pattern.compile("Product Catalog"));
    System.out.println("Blocked " + blocked.size() + " analytics request(s)");
}

4. Run the tests

mvn test

To run in parallel, configure the Surefire plugin in pom.xml:

<plugin>
    <groupId>org.apache.maven.plugins</groupId>
    <artifactId>maven-surefire-plugin</artifactId>
    <configuration>
        <parallel>methods</parallel>
        <threadCount>4</threadCount>
    </configuration>
</plugin>

5. Cross-browser testing

Use @ParameterizedTest with @MethodSource to run the same test method against each browser endpoint. Each invocation opens its own Playwright instance so browser types don't share state:

import org.junit.jupiter.params.ParameterizedTest;
import org.junit.jupiter.params.provider.*;
import java.util.stream.Stream;

static Stream<Arguments> browserEndpoints() {
    String token = "YOUR_API_TOKEN_HERE";
    String base = "wss://production-sfo.browserless.io";
    return Stream.of(
        Arguments.of("chromium", base + "/chromium/playwright?token=" + token),
        Arguments.of("firefox",  base + "/firefox/playwright?token="  + token),
        Arguments.of("webkit",   base + "/webkit/playwright?token="   + token)
    );
}

@ParameterizedTest(name = "{0}")
@MethodSource("browserEndpoints")
void pageHasCorrectTitle(String browserName, String wsEndpoint) {
    try (Playwright pw = Playwright.create()) {
        Browser b = pw.chromium().connect(wsEndpoint);
        Page p = b.newPage();
        p.navigate("https://scraping-sandbox.netlify.app/products");
        assertThat(p).hasTitle(Pattern.compile("Product Catalog"));
    }
}

6. Debug with Trace Viewer

Start and stop tracing in @BeforeEach / @AfterEach so each test gets its own trace file:

import java.nio.file.Paths;
import org.junit.jupiter.api.TestInfo;

@BeforeEach
void createContextAndPage() {
    context = browser.newContext();
    context.tracing().start(new Tracing.StartOptions()
        .setScreenshots(true).setSnapshots(true));
    page = context.newPage();
}

@AfterEach
void closeContext(TestInfo testInfo) {
    context.tracing().stop(new Tracing.StopOptions()
        .setPath(Paths.get("trace-" + testInfo.getDisplayName() + ".zip")));
    context.close();
}

Open the trace:

mvn exec:java -e -Dexec.mainClass=com.microsoft.playwright.CLI -Dexec.args="show-trace trace.zip"

7. Check the output

[INFO] Tests run: 4, Failures: 0, Errors: 0, Skipped: 0

View Full Code on GitHub

1. Install dependencies

dotnet add package Microsoft.Playwright.NUnit

2. Set up the test class

Connect to Browserless once in [OneTimeSetUp] so the WebSocket handshake only happens once per suite. Open a fresh context per test in [SetUp]. Separate contexts give each test isolated cookies, storage, and network state, so tests can't bleed state into each other:

using Microsoft.Playwright;
using NUnit.Framework;
using System.Collections.Generic;
using System.Threading.Tasks;

[Parallelizable(ParallelScope.Self)]
[TestFixture]
public class ExampleTests
{
    private IPlaywright _playwright;
    private IBrowser _browser;
    private IBrowserContext _context;
    private IPage _page;

    [OneTimeSetUp]
    public async Task OneTimeSetUp()
    {
        _playwright = await Playwright.CreateAsync();
        _browser = await _playwright.Chromium.ConnectAsync(
            "wss://production-sfo.browserless.io/chromium/playwright?token=YOUR_API_TOKEN_HERE"
        );
    }

    [SetUp]
    public async Task SetUp()
    {
        _context = await _browser.NewContextAsync();
        _page = await _context.NewPageAsync();
    }

    [TearDown]
    public async Task TearDown()
    {
        await _context.CloseAsync();
    }

    [OneTimeTearDown]
    public async Task OneTimeTearDown()
    {
        await _playwright.DisposeAsync();
    }
}

3. Write tests

using System.Text.RegularExpressions;

[Test]
public async Task PageHasCorrectTitle()
{
    await _page.GotoAsync("https://scraping-sandbox.netlify.app/products");
    await Assertions.Expect(_page).ToHaveTitleAsync(new Regex("Product Catalog"));
}

[Test]
public async Task HeadingIsVisible()
{
    await _page.GotoAsync("https://scraping-sandbox.netlify.app/products");
    var heading = _page.Locator("h1");
    await Assertions.Expect(heading).ToBeVisibleAsync();
    await Assertions.Expect(heading).ToHaveTextAsync("Product Catalog");
}

[Test]
public async Task PageHasLink()
{
    await _page.GotoAsync("https://scraping-sandbox.netlify.app/products");
    await Assertions.Expect(_page.Locator("a[href]").First).ToBeVisibleAsync();
}

[Test]
public async Task PageLoadsWithoutAnalytics()
{
    var blocked = new List<string>();
    await _page.RouteAsync("**/analytics/**", async route =>
    {
        blocked.Add(route.Request.Url);
        await route.AbortAsync();
    });
    await _page.GotoAsync("https://scraping-sandbox.netlify.app/products");
    await Assertions.Expect(_page).ToHaveTitleAsync(new Regex("Product Catalog"));
    Console.WriteLine($"Blocked {blocked.Count} analytics request(s)");
}

4. Run the tests

dotnet test

To configure parallelism, add a .runsettings file:

<RunSettings>
    <RunConfiguration>
        <MaxCpuCount>4</MaxCpuCount>
    </RunConfiguration>
</RunSettings>

Then run with dotnet test --settings .runsettings.

5. Cross-browser testing

Use NUnit's [TestCaseSource] to feed each browser endpoint as a separate test case. Each case creates its own Playwright instance so browser types don't share state:

private static IEnumerable<TestCaseData> BrowserEndpoints()
{
    var token = "YOUR_API_TOKEN_HERE";
    var @base = "wss://production-sfo.browserless.io";
    yield return new TestCaseData("chromium", $"{@base}/chromium/playwright?token={token}");
    yield return new TestCaseData("firefox",  $"{@base}/firefox/playwright?token={token}");
    yield return new TestCaseData("webkit",   $"{@base}/webkit/playwright?token={token}");
}

[Test, TestCaseSource(nameof(BrowserEndpoints))]
public async Task PageHasCorrectTitle(string browserName, string wsEndpoint)
{
    var playwright = await Playwright.CreateAsync();
    var browser = await playwright.Chromium.ConnectAsync(wsEndpoint);
    var page = await browser.NewPageAsync();
    await page.GotoAsync("https://scraping-sandbox.netlify.app/products");
    await Assertions.Expect(page).ToHaveTitleAsync(new Regex("Product Catalog"));
    await playwright.DisposeAsync();
}

6. Debug with Trace Viewer

Start tracing in [SetUp] and stop it in [TearDown] so each test gets its own trace file:

[SetUp]
public async Task SetUp()
{
    _context = await _browser.NewContextAsync();
    await _context.Tracing.StartAsync(new TracingStartOptions
    {
        Screenshots = true,
        Snapshots = true,
    });
    _page = await _context.NewPageAsync();
}

[TearDown]
public async Task TearDown()
{
    await _context.Tracing.StopAsync(new TracingStopOptions
    {
        Path = $"trace-{TestContext.CurrentContext.Test.Name}.zip",
    });
    await _context.CloseAsync();
}

Open the trace:

pwsh bin/Debug/net8.0/playwright.ps1 show-trace trace.zip

7. Check the output

Test Run Successful.
Total tests: 4
     Passed: 4
 Total time: 2.5000 Seconds

View Full Code on GitHub

1. Write the mutation

Navigate to the page and select the content you want to verify. BQL returns typed structured data, so you write your assertions in your own application code after the response arrives. No assertion library runs inside Browserless:

mutation E2ECheck {
  goto(url: "https://scraping-sandbox.netlify.app/products", waitUntil: domContentLoaded) {
    status
  }

  title: mapSelector(selector: "title") {
    innerText
  }

  heading: mapSelector(selector: "h1") {
    innerText
  }

  links: mapSelector(selector: "a[href]") {
    href: attribute(name: "href") {
      value
    }
  }
}

2. Run it

Paste into the BQL IDE and click Run.

3. Check the output

{
  "data": {
    "goto": { "status": 200 },
    "title": [{ "innerText": "Product Catalog" }],
    "heading": [{ "innerText": "Product Catalog" }],
    "links": [{ "href": { "value": "https://scraping-sandbox.netlify.app/contact-us" } }]
  }
}

Assert against the values in your application code. Check goto.status for the HTTP status code, innerText fields for content, array lengths for element counts, and attribute fields for link targets.

Next steps

Run Concurrent Browser Sessions

run tests in parallel across multiple browsers

Log In and Reuse Sessions

authenticate once and reuse state across test runs

Disconnect and Reconnect to a Browser

pause and resume long-running test sessions

Steps​

Next steps​

Run Concurrent Browser Sessions

Log In and Reuse Sessions

Disconnect and Reconnect to a Browser

Steps

Next steps