QA Proof of Work

Automation Exercise

End-to-End QA Automation & Continuous Testing

An independent, self-initiated QA project, not client work, run against automationexercise.com, a public practice store. It combines manual testing, REST API validation, Playwright UI automation, Page Object Model architecture, debugging and GitHub Actions CI.

View Source Code View CI Runs

Last Recorded CI Result (29 Aug 2026) Passed

26/26 passed

26 test definitions (16 API, 10 UI) passed in the GitHub Actions run of 29 August 2026 (commit 789d413), with CI retries enabled. This is a historical result, not a current execution status.

0 Failed

Chromium Browser project

0 Skipped

API + UI Coverage layers

26 Automated Tests

API + UI Testing Layers

Playwright Automation Framework

POM Framework Architecture

CI GitHub Actions

The Challenge

How do you validate an e-commerce application beyond basic UI checks?

Rather than treating automation as a collection of isolated browser scripts, I approached the application through multiple QA layers.

The project combines test planning, manual and exploratory validation, API testing, UI automation, regression execution and release assessment.

The goal was to build a maintainable automation suite that could provide repeatable evidence rather than simply demonstrate that a few browser actions could be automated.

Plan

Define scope, risks and test scenarios.

Validate

Perform manual and exploratory testing.

API

Validate REST endpoints and negative cases.

Automate

Build Playwright UI regression coverage.

Integrate

Execute automated tests through GitHub Actions.

Assess

Review execution results and quality findings.

Test Coverage

What I tested

The automated suite covers the application's core customer journeys and selected API behaviours.

Authentication

Registration
Duplicate registration
Valid login
Invalid credentials
Account deletion

Products

Product listing
Product search
Brands
API validation

Shopping Cart

Single product
Multiple products
Product removal
Quantity validation
Guest checkout

Checkout

Order review
Product verification
Order comments
Payment flow
Order confirmation

API

Account lifecycle
Authentication
Product search
Brands
Negative scenarios

Framework Architecture

Built for maintainability, not just execution.

The automation framework separates test intent from page interaction logic and supports both UI and API testing.

Spec files call page objects (UI) or API client classes (API). Structure read from the repository.

Page Object Model

Page interaction logic is separated from individual test scenarios to improve maintainability.

Reusable Utilities

Common operations are centralized instead of being repeatedly implemented inside individual tests.

Authentication State

Tests can begin from an authenticated state where the scenario requires it.

API + UI Layers

The suite validates behaviour beyond the browser layer by including direct API testing.

A page object, from the repository

export class LoginPage {
  constructor(page) {
    this.page = page;
    // ...
    // Login section
    this.loginEmailInput = page.locator('input[data-qa="login-email"]');
    this.loginPasswordInput = page.locator('input[data-qa="login-password"]');
    this.loginButton = page.locator('button[data-qa="login-button"]');
  // ...
  async login(email, password) {
    await this.loginEmailInput.fill(email);
    await this.loginPasswordInput.fill(password);
    await this.loginButton.click();
  }
}

LoginPage: locators live in the constructor, tests call login(). Condensed. pages/LoginPage.js, lines 1–3, 10–13, 30–35 @ 789d413

Continuous Testing

Automated regression through GitHub Actions

The project uses GitHub Actions to execute API and UI test suites in CI. The workflow is configured to upload Playwright reports as artifacts (14-day retention; none are currently available, checked 12 October 2026).

From .github/workflows/playwright.yml. The two jobs run separately.

What the pipeline proves

The project is not dependent on a local machine to demonstrate that the automation works. The test suite is executable through the CI workflow.

Playwright dependencies installed in CI
Chromium installed for UI execution
API and UI suites executed independently
Playwright reports configured as workflow artifacts (14-day retention, now expired)
Workflow triggered by repository changes

Inspect GitHub Actions

Engineering & Debugging

Automation isn't useful if it isn't stable.

Checkout automation required investigation and stabilization.

During automation development, checkout scenarios exposed synchronization and cart-state problems. Instead of simply increasing timeouts, I used the available Playwright evidence to investigate the failure path and refine the automation.

01 · FAILURE Test exposes unexpected behaviour.

02 · EVIDENCE Screenshot, video and trace reviewed.

03 · ANALYSIS Locator and synchronization behaviour investigated.

04 · REFINEMENT Page-object interaction and waiting strategy refined.

05 · VERIFY Suite re-run to confirm the regression result. A local no-retry full UI run on 12 October 2026 had 2 failures out of 10 (a checkout assertion and a 60-second login timeout); a repeat of those two tests three times and a second full run passed. Treat the UI suite as one that can fail transiently.

QA Outcome

Evidence over claims.

100% Pass rate of the last recorded CI run (29 Aug 2026); a historical result, not a current status

26 Automated tests executed in the last recorded CI run (29 Aug 2026)

0 Critical defects

0 High-severity defects

2 Minor UI issues documented

QA finding

The documented QA assessment also records minor UI issues and a third-party ad injection behaviour. These findings are presented as observations from the tested environment rather than being described as application-critical production defects.

Proof & Documentation

Inspect the work.

The project is backed by source code, QA documentation and CI execution evidence.

Source Code

Explore the Playwright framework, page objects, API tests and utilities.

Open Repository

Test Plan

Review the project scope, testing approach, risks and planned validation.

View Test Plan

Manual Test Cases

Review the manually designed test scenarios supporting the project.

View Test Cases

Execution Report

Review the documented execution results and QA assessment.

View Report

CI Pipeline

Review the GitHub Actions run history. Job logs require a GitHub sign-in.

View CI Runs

Latest Passing Run

Open the GitHub Actions run of 29 August 2026 (commit 789d413), where the API and UI jobs both passed. Job logs require a GitHub sign-in, and Playwright reports are not committed to the repository.

View Passing Run

Technology

Tools used in the project

Playwright JavaScript Node.js REST API Testing Page Object Model GitHub Actions Chromium Manual Testing Regression Testing

What This Demonstrates

More than a Playwright script.

QA Thinking

Test planning and validation were considered before building the automated regression layer.

Automation Engineering

The framework uses page objects and reusable components instead of placing every interaction directly inside test files.

API Testing

Application behaviour is validated at both the API and browser layers.

Debugging

Playwright screenshots, videos and traces are used as evidence when investigating failures.

CI Engineering

Automated tests can execute through GitHub Actions instead of relying only on local runs.

Quality Communication

Test planning, execution evidence and QA findings are documented alongside the code.

Don't just take my word for it. Inspect the implementation.

The source code, CI workflow and QA documentation are available for review.

View GitHub Verify CI Back to Proof of Work