This repo is a public, production‑style example of how I design and build test automation: UI, API, performance and CI/CD. Real client work is under NDA, but the structure, patterns and documentation here match what I use on real projects.

Look at these first:

1. How I structure automation

   • docs/architecture.md
   • docs/test_strategy.md

2. How tests map to requirements

   • docs/traceability_matrix.md
   • docs/test_cases_azercell_login.md

3. Concrete UI automation (Selenium + pytest)

   • Tests: autotests/test_azercell_login_page.py, autotests/test_smoke.py
   • Page Objects: pages/azercell_login_page.py, pages/base_page.py

4. API + performance examples

   • API: postman/collections/restful-booker.postman_collection.json
   • Performance: performance/restful-booker-smoke.js

5. Process / onboarding

   • docs/onboarding_guide_for_new_qa.md

If you skim only a few files, those will give you an accurate view of how I work.

• Web UI
  Azercell “Kabinetim” login flow automated with:
  • Python, pytest
  • Selenium WebDriver
  • Page Object Model (POM)
• API
  Restful Booker API tested with:
  • Postman collections
  • Newman CLI (HTML + JUnit reports)
• Performance smoke
  Light k6 script against Restful Booker to show how I wire basic performance checks into CI.

• GitHub Actions workflow: .github/workflows/ci.yml
• On every push and pull request:
  • lint job runs ruff + black as a quality gate
  • ui-tests job runs Selenium smoke tests in headless Chrome
  • api-tests job runs the Postman collection via Newman
• Nightly schedule:
  • UI: smoke + regression pack
  • API: full collection
  • Performance: k6 smoke test
• All jobs publish artifacts (JUnit XML, screenshots, HTML, Newman reports) to make debugging straightforward.

• Focus on small, reliable smoke suites that protect the most important flows:
  • Azercell home page reachable
  • Login button navigates correctly
  • Phone input works and validates
• Deeper regression tests:
  • Additional navigation rules
  • Password‑change flow (when available)
  • Format variations and validation behaviour

• Clear separation of concerns:
  • Tests live in autotests/
  • Page Objects live in pages/
  • Utilities in utils/
  • Configuration and fixtures in conftest.py
• Page Objects hide Selenium details behind readable methods like:
  • open_home_page()
  • click_login_button()
  • enter_phone_number()
  • submit_phone_number()
• Robustness features:
  • Multiple locators and fallbacks for key elements
  • Centralised BasePage.click() with retries and JS fallback
  • Window/tab switching when the site opens login in a new window
  • Automatic screenshots and HTML capture on test failure

• Linters and formatters:
  • ruff for static analysis
  • black for formatting
• Pre‑commit configuration:
  • .pre-commit-config.yaml to keep the repo consistently formatted
• Documentation:
  • Test strategy, architecture, onboarding, test cases, traceability
  • Lightweight ADRs in docs/adr/ for architectural decisions
• Flaky tests:
  • Explicit policy using pytest-rerunfailures
  • Only marked flaky when the external system is unstable; root cause is documented and targeted for fixing

• Language & test runner
  • Python 3.11+
  • pytest
• Web UI automation
  • Selenium WebDriver
  • Chrome / Chromium (headless in CI)
  • Page Object Model pattern
• API testing
  • Postman collections
  • Newman CLI (cli, html, junit reporters)
• Performance testing
  • Grafana k6 (local CLI and GitHub Action)
• CI/CD
  • GitHub Actions (.github/workflows/ci.yml)
• Support tooling
  • python-dotenv for .env handling
  • pytest-xdist, pytest-timeout, pytest-rerunfailures

Reflecting what you see in the IDE screenshots:

.github/
  workflows/
    ci.yml                   # GitHub Actions pipeline (lint, UI, API, perf)

autotests/
  README_test_azercell_login.md
  test_azercell_login_page.py   # Main Azercell UI flows (smoke + regression)
  test_smoke.py                 # Minimal smoke test (home → login)

docs/
  adr/
    0001-use-pytest-selenium-pom.md
    0002-ci-gating-smoke-on-pr.md
  architecture.md
  onboarding_guide_for_new_qa.md
  test_cases_azercell_login.md
  test_strategy.md
  traceability_matrix.md        # Test cases → automated tests

pages/
  azercell_login_page.py        # Azercell login Page Object
  base_page.py                  # Shared base Page Object (click, open, windows)
  login_page.py                 # Generic placeholder LoginPage
  __init__.py

postman/
  data/
  environments/
    restful-booker.postman_environment.json
  collections/
    restful-booker.postman_collection.json
  newman/
    run-restful-booker.sh
  README.md

performance/
  restful-booker-smoke.js       # k6 API smoke test

reports/                        # Created at runtime (.gitkeep only)
  downloads/
  screenshots/

utils/
  phone.py                      # normalize_phone_number() helper

conftest.py                     # pytest fixtures, WebDriver, screenshots on fail
pytest.ini                      # markers, default pytest config
requirements.txt                # Python dependencies
.pre-commit-config.yaml         # black + ruff hooks
.gitignore
LICENSE
README.md                       # This file

• Python 3.11+
• Node.js 18+ (for Newman)
• Google Chrome / Chromium installed

Clone the repo and create a virtual environment:

git clone https://github.com/<your-username>/<this-repo>.git
cd <this-repo>

python -m venv .venv
source .venv/bin/activate  # Windows: .venv\Scripts\activate

Install dependencies:

pip install --upgrade pip
pip install -r requirements.txt

Optional (for API tests):

npm install -g newman newman-reporter-html newman-reporter-junitfull

Basic environment configuration:

export TEST_ENV="local"
export BASE_URL="https://www.azercell.com/az"
export HEADLESS="1"              # use "0" to see the browser
export WAIT_TIMEOUT="15"
export PAGE_LOAD_TIMEOUT="45"

The Azercell login tests can use a real phone number. For safety, the placeholder 5XXXXXXXXX is treated as “no real number” and deeper flows are automatically skipped.

# Optional – only if you have a real test number
export AZERCELL_PHONE="5XXXXXXXXX"  # format 5XXXXXXXXX, no leading 0

Run the smoke suite (same idea as PR checks):

pytest autotests -m smoke -v

Run smoke + regression (broader coverage):

pytest autotests -m "smoke or regression" -v

On failure, screenshots and page source are stored under:

reports/screenshots/<test_nodeid>_<timestamp>.png
reports/screenshots/<test_nodeid>_<timestamp>.html

newman run postman/collections/restful-booker.postman_collection.json \
  -e postman/environments/restful-booker.postman_environment.json \
  --reporters cli,junit,html \
  --reporter-junit-export reports/api/postman-junit.xml \
  --reporter-html-export reports/api/postman-results.html

These tests cover basic happy paths and negative flows of the Restful Booker demo API.

export BASE_URL="https://restful-booker.herokuapp.com"
k6 run performance/restful-booker-smoke.js

This is a small load test that exercises /ping with thresholds on latency and error rate, meant to show how performance checks can be integrated into CI rather than to stress the system.

Workflow: .github/workflows/ci.yml

• Triggers:
  • push to main
  • pull_request targeting main
  • workflow_dispatch (manual)
  • Nightly schedule (0 0 * * *)

Jobs:

1. lint

   • Installs ruff and black.
   • Runs:
     • ruff check pages autotests
     • black --check pages autotests

2. ui-tests

   • Needs lint.
   • Installs Chrome on Ubuntu runner.
   • Sets:
     • BASE_URL=https://www.azercell.com/en
     • HEADLESS=1
   • Injects phone via PHONE_NUMBER=${{ secrets.AZERCELL_PHONE_NUMBER }} when configured.
   • Runs:
     • On push/PR: pytest autotests -m "smoke" -v
     • On nightly schedule: pytest autotests -m "smoke or regression" -v
   • Uploads:
     • reports/ui/ui-junit.xml
     • Screenshots and page source files.

3. api-tests

   • Needs lint.
   • Sets up Node and Newman.
   • Ensures a Restful Booker environment JSON exists.
   • Runs the Postman collection and uploads:
     • reports/api/postman-junit.xml
     • reports/api/postman-results.html.

4. perf-smoke

   • Needs api-tests.
   • Runs only on schedule or manual dispatch.
   • Uses official k6 GitHub Actions to execute performance/restful-booker-smoke.js.

This pipeline shows how I typically gate changes with fast checks on every PR and push, plus broader regression and performance checks on a schedule.

In production environments I have:

• Worked with mobile (iOS/Android) automation on BrowserStack
• Built BI / analytics data checks with SQL and API orchestration
• Integrated with TestRail, Jira and release processes
• Owned regression planning and CI gating strategies

Those repos are private, but this project mirrors my approach: structured automation, strong documentation, and CI as the default way to run tests.

• Email: [email protected]
• LinkedIn: https://www.linkedin.com/in/huseyn-gasimov-
• Location: Baku, Azerbaijan