Running Tests

Three ways to run, in increasing order of control: Make, npm scripts, raw npx.

Quick reference

make test               # everything
make test-smoke         # 1 user — fastest sanity check
make test-core          # 2 users — adds main error path
make test-full          # all 6 users — full regression
make test-ui            # Playwright UI mode (interactive)
make test-debug         # Inspector — step through manually
make report             # reopen the last HTML report

make help lists every target.

npm scripts

For finer control. Defined in package.json:

Command	What it does
npm test	Run all tests
npm run test:chromium	Chromium only
npm run test:firefox / :webkit	Single browser
npm run test:headed / :headless	Force visibility mode
npm run test:smoke / :core / :full	Profile-filtered
npm run test:login	Login suite only
npm run test:ci	CI mode (headless, no auto-open)
npm run test:debug	Playwright Inspector (PWDEBUG=1)
npm run test:ui	Playwright UI mode (interactive)
npm run test:no-retry	Skip retries (debugging flakes)
npm run report	Reopen last HTML report
npm run trace	Open a trace file

Raw npx playwright

When no script fits:

# By test title substring
npx playwright test -g "standard_user" --project=chromium

# By file:line
npx playwright test tests/saucedemo/login.spec.ts:42

# Run a cookbook file directly
npx playwright test tests/saucedemo/network.cookbook.ts --project=chromium

# Repeat to expose flakes
npx playwright test -g "name" --repeat-each=20 --retries=0

# Generate a test by recording your clicks
npx playwright codegen https://www.saucedemo.com

Full CLI flag reference: CLI Reference.

Test artifacts

Every run writes to test-results/<test-name>/:

File	How to open
trace.zip	make trace FILE=test-results/<name>/trace.zip or drop into trace.playwright.dev
video.webm	Any video player or your browser
*.png	Screenshot at test completion

The playwright-report/ folder contains the HTML report with all of the above embedded. make report reopens the last one.

CI runs

Every push and PR runs the suite via GitHub Actions. The HTML report is uploaded as an artifact and mirrored to this docs site at /test-report/ after every push to main.

CI integration deep-dive: CI / CD Reference.

What "passes" really means

This project's config:

• retries: 3 — a test is retried up to 3× before being marked failed
• timeout: 60_000 — each test has 60s to complete
• actionTimeout: 15_000 — each action (click, fill) has 15s

If a test passes after retries, the HTML report flags it as flaky — not the same as green. Flaky tests are worth fixing; see Debugging → Diagnosing flaky tests.