Capstone Project — Building a Complete Test Automation Framework from Scratch

You have spent 79 chapters learning testing fundamentals, manual techniques, API testing, Selenium, Playwright, Cypress, CI/CD, security, accessibility, BDD, and SDET skills. This capstone project brings everything together into a single, portfolio-ready deliverable: a complete test automation framework for a real web application. The project follows the five-layer architecture from Chapter 42 and incorporates patterns from every part of the series.

Capstone Project — Scope, Architecture and Planning #

The capstone framework targets a public practice application (SauceDemo, The Internet, or similar) and demonstrates every skill from the series.

# Capstone project scope

CAPSTONE_SCOPE = {
    "target_app": "SauceDemo (https://www.saucedemo.com) or equivalent practice site",
    "tool_choice": "Selenium (Python), Cypress (TypeScript), or Playwright (TypeScript/Python)",
    "deliverables": [
        "Complete framework with five-layer architecture",
        "15-20 E2E test cases covering critical user journeys",
        "5-10 API test cases with request/response validation",
        "Page Object Model with BasePage and 4-6 page objects",
        "Data-driven tests with external fixture files",
        "Custom commands/utilities for reusable actions",
        "CI/CD pipeline (GitHub Actions) with parallel execution",
        "Cross-browser configuration (Chrome + Firefox minimum)",
        "HTML test report with failure screenshots",
        "Security test cases (XSS, access control)",
        "Accessibility checks (axe-core integration)",
        "README documentation with setup and run instructions",
    ],
}

# Project structure
PROJECT_STRUCTURE = """
capstone-qa-framework/
  config/
    settings.py (or cypress.config.ts)     # Layer 1: Configuration
    browsers.py                             # Browser-specific options
  pages/ (or cypress/support/)
    base_page.py                            # Layer 3: BasePage with shared methods
    login_page.py                           # LoginPage — locators + actions
    inventory_page.py                       # InventoryPage
    cart_page.py                            # CartPage
    checkout_page.py                        # CheckoutPage
  tests/ (or cypress/e2e/)
    test_login.py                           # Layer 4: Login test cases
    test_inventory.py                       # Product browsing tests
    test_cart.py                            # Cart management tests
    test_checkout.py                        # End-to-end checkout flow
    test_api_products.py                    # API validation tests
    test_security.py                        # XSS + access control tests
  utils/
    driver_factory.py                       # Layer 2: Driver management
    data_builder.py                         # Layer 5: Test data builders
    reporter.py                             # Screenshot + logging on failure
  test_data/
    users.json                              # Login credentials fixture
    products.json                           # Product data fixture
  .github/workflows/
    ci.yml                                  # CI/CD pipeline
  reports/                                  # Generated (gitignored)
  requirements.txt (or package.json)
  README.md                                 # Setup + run instructions
"""

# Planning checklist
PLANNING_CHECKLIST = [
    "1. Choose your tool: Selenium (Python), Cypress (TS), or Playwright (TS/Python)",
    "2. Set up the project: initialise repo, install dependencies, create folder structure",
    "3. Implement Layer 1 (Config): environment variables, base URL, timeouts, browser selection",
    "4. Implement Layer 2 (Driver): DriverFactory or browser fixture with local/Grid support",
    "5. Implement Layer 3 (Pages): BasePage + 4-6 page objects with locators and actions",
    "6. Implement Layer 4 (Tests): 15-20 test cases across login, inventory, cart, checkout",
    "7. Implement Layer 5 (Utils): data builders, reporters, screenshot on failure",
    "8. Add API tests: 5-10 tests using cy.request or requests library",
    "9. Add security tests: XSS payloads, access control checks",
    "10. Add accessibility: axe-core integration in functional tests",
    "11. Configure CI/CD: GitHub Actions with Chrome + Firefox, parallel execution",
    "12. Write README: setup instructions, run commands, architecture description",
    "13. Review and refactor: apply SOLID principles, clean code, consistent naming",
]

print("Capstone Project Plan:")
print(PROJECT_STRUCTURE)
print("\nPlanning Checklist:")
for item in PLANNING_CHECKLIST:
    print(f"  {item}")

Common Mistakes #

Mistake 1 — Starting with tests instead of architecture #

❌ Wrong: Writing 20 test files with duplicated setup, no page objects, and hardcoded data — then trying to refactor into a framework.

✅ Correct: Setting up the five-layer structure first (config, driver, pages, tests, utils), implementing BasePage and one page object, writing one test, then expanding from that solid foundation.

Mistake 2 — Not committing incrementally to Git #

❌ Wrong: Building the entire framework locally and pushing one massive commit at the end.

✅ Correct: Committing after each milestone: “Add BasePage and LoginPage,” “Add login tests,” “Add CI pipeline,” “Add API tests.” The Git history demonstrates your development process and thinking.