testland/cloud-grid-e2e

Author and run E2E tests on a cloud browser grid - BrowserStack Automate, Sauce Labs, or LambdaTest. All three follow one pattern: username + access-key env vars, a W3C WebDriver hub URL, a vendor options dict inside the capabilities (bstack:options / sauce:options / LT:Options), a local tunnel binary for internal apps, session pass/fail reporting, and a CI matrix throttled to the plan's parallel-session limit. Worked example uses BrowserStack; per-vendor deltas live in references/. Use for cross-browser regression on real devices + browsers beyond the engines bundled on the local machine - distinct from a local matrix runner and from self-hosted Selenium Grid.

Quality

90%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Sauce Labs - vendor detail

Name: testland/cloud-grid-e2e
Rating: 72 (1 reviews)
Author: testland

Deep reference for cloud-grid-e2e. Sauce Labs is one of the original cloud-grid providers, with a W3C-compliant WebDriver endpoint covering desktop + mobile browser combinations and strong Cypress / Playwright / Appium support. Per docs.saucelabs.com/dev/test-configuration-options.

Also worth knowing: running the same suite on Sauce + a second grid is a practical way to catch grid-specific flakes.

Auth + hub URLs (regional)

export SAUCE_USERNAME="oauth-...-..."
export SAUCE_ACCESS_KEY="<access-key-from-user-settings>"

US-West:    https://ondemand.us-west-1.saucelabs.com:443/wd/hub
US-East:    https://ondemand.us-east-4.saucelabs.com:443/wd/hub
EU-Central: https://ondemand.eu-central-1.saucelabs.com:443/wd/hub

Pick the region closest to your CI runner for lower latency. Each region has its own device matrix + availability; mixing regions in one run adds flake.

Capabilities (W3C)

{
  "browserName": "chrome",
  "browserVersion": "latest",
  "platformName": "Windows 11",
  "sauce:options": {
    "build": "PR-1234",
    "name": "Login flow on Chrome Windows",
    "username": "$SAUCE_USERNAME",
    "accessKey": "$SAUCE_ACCESS_KEY",
    "screenResolution": "1920x1080",
    "tunnelName": "my-internal-tunnel",
    "extendedDebugging": true,
    "capturePerformance": true,
    "recordVideo": true,
    "recordScreenshots": true,
    "tags": ["smoke", "e2e", "auth"]
  }
}

`sauce:options` field	Purpose
`username` / `accessKey`	Credentials (can also be in URL)
`build`	Group sessions by CI build
`name`	Session label in dashboard
`screenResolution`	Default 1024x768; common: 1920x1080
`tunnelName`	Sauce Connect tunnel reference (preferred over deprecated `tunnelIdentifier`)
`extendedDebugging`	Enable HAR + console + Selenium logs
`capturePerformance`	Browser performance metrics
`recordVideo` / `recordScreenshots`	Session capture
`tags`	Free-form tags for filtering

browserVersion accepts "latest", "latest-1", etc. - version-relative pinning works across release cycles.

Python example

import os
from selenium import webdriver

options = webdriver.FirefoxOptions()
options.browser_version = "latest"
options.platform_name = "Windows 11"

sauce_options = {
    "build": os.environ.get("BUILD_TAG", "local"),
    "name": "Checkout flow Firefox",
    "username": os.environ["SAUCE_USERNAME"],
    "accessKey": os.environ["SAUCE_ACCESS_KEY"],
    "screenResolution": "1920x1080",
    "extendedDebugging": True,
}

# Vendor caps must be set on Options BEFORE Remote() ([Selenium options]).
options.set_capability("sauce:options", sauce_options)

driver = webdriver.Remote(
    command_executor="https://ondemand.us-west-1.saucelabs.com:443/wd/hub",
    options=options,
)

driver.get("https://example.com")
# test...
driver.quit()

Session status

driver.execute_script("sauce:job-result=" + ("passed" if not failed else "failed"))

Or via REST API: PUT /rest/v1/{username}/jobs/{session_id}.

Sauce Connect Proxy

Per docs.saucelabs.com/secure-connections/sauce-connect-5:

# Download Sauce Connect 5 from saucelabs.com
./sc \
  --username $SAUCE_USERNAME \
  --access-key $SAUCE_ACCESS_KEY \
  --tunnel-name "my-internal-tunnel" \
  --region us-west-1

Then set sauce:options.tunnelName: "my-internal-tunnel" in capabilities. Tunnel cleans up on Ctrl+C. For ephemeral CI: spawn → wait-for-ready → run tests → terminate. SC tunnel setup adds 10-30s to test start.

Parsing results

Session reports include: video (always - recordVideo: true default), network HAR (if extendedDebugging), browser console logs, Selenium logs, per-command screenshots (if recordScreenshots), performance metrics (if capturePerformance). Retrieve via REST (docs.saucelabs.com/dev/api):

curl -u "$SAUCE_USERNAME:$SAUCE_ACCESS_KEY" \
  "https://api.us-west-1.saucelabs.com/rest/v1/$SAUCE_USERNAME/jobs/<session-id>"

CI integration

on: pull_request
jobs:
  sauce:
    runs-on: ubuntu-latest
    strategy:
      matrix:
        browser:
          - { name: chrome, version: latest, platform: "Windows 11" }
          - { name: safari, version: "17", platform: "macOS 14" }
          - { name: edge, version: latest, platform: "Windows 11" }
    steps:
      - uses: actions/checkout@v6
      - name: Run on Sauce Labs
        env:
          SAUCE_USERNAME: ${{ secrets.SAUCE_USERNAME }}
          SAUCE_ACCESS_KEY: ${{ secrets.SAUCE_ACCESS_KEY }}
          SAUCE_BROWSER: ${{ matrix.browser.name }}
          SAUCE_VERSION: ${{ matrix.browser.version }}
          SAUCE_PLATFORM: ${{ matrix.browser.platform }}
          BUILD_TAG: pr-${{ github.event.pull_request.number }}
        run: pytest tests/e2e/ --sauce

Vendor-specific anti-patterns

Anti-pattern	Why it fails	Fix
Hardcoded region	Cross-region latency adds 100ms+ per command	Match region to CI runner location
`tunnelIdentifier` (deprecated)	Newer SC versions emit warnings	Use `tunnelName`
Mixed regions in one test run	Increases flake	Pick one region per run
`recordVideo: false` to "save money"	Failed-session debugging hard	Keep video for failed sessions at minimum