saucedemo-qa

Manual and automated QA of SauceDemo, a demo e-commerce application. This project shows the full path from exploratory testing to a written bug report to an automated regression test built from that report.

What this project demonstrates

• Exploratory testing of the main purchase flow (login, inventory, cart, checkout), using a reference account to establish correct behavior.
• Professional bug reports — documented defects with reproduction steps, expected vs. actual results, and severity ratings (see bug-reports.md).
• Bug triage discipline — suspected issues were validated against a known-good reference account before being reported, and false positives were ruled out instead of filed.
• Automated regression testing — a Playwright test that encodes correct behavior; it passes against the working account and catches a real defect when run against the broken one.

Approach

Two SauceDemo accounts were used deliberately: standard_user as a baseline expected to work correctly, and problem_user as the account under test. Comparing the two made it possible to tell genuine defects apart from expected behavior — and to discard false alarms instead of filing incorrect reports.

How it works

flowchart TD
    site["SauceDemo (live site)"]

    subgraph manual["Manual exploratory testing"]
        acc1["standard_user — baseline"]
        acc2["problem_user — under test"]
        acc1 --> compare["compare behavior"]
        acc2 --> compare
        compare --> triage["confirm real defects, rule out false positives"]
        triage --> reports["bug reports (docs/bug-reports.md)"]
    end

    site --> acc1
    site --> acc2
    reports --> pick["one defect: broken sort"]

    subgraph auto["Automated regression"]
        pick --> test["tests/test_sorting.py — asserts correct sort order"]
        test --> passes["passes for standard_user"]
        test --> fails["fails for problem_user — catches the defect"]
    end

    push["git push"] --> ci["GitHub Actions (CI)"]
    ci --> test

Loading

From bug report to regression test

The automated test was built directly from one of the defects found during exploratory testing — the broken sort functionality. Here is the report it is based on:

---

Defect — Inventory: sort dropdown options are unresponsive

Environment: saucedemo.com · user: problem_user · Chrome

Steps to reproduce:

1. Log in as problem_user / secret_sauce.
2. Click the sort dropdown menu at the top right to open it.
3. Attempt to select a different sorting option (e.g., "Price (high to low)").

Expected result: The selected option is applied and the products are reordered based on the new criteria.

Actual result: The dropdown opens, but clicking a sorting option triggers no action. The selection is ignored and the product order does not change.

Severity: Medium — restricts navigation and user experience, though it does not outright block the purchasing path.

---

From this finding, tests/test_sorting.py was written to encode the correct behavior: after selecting "Price (low to high)", the product prices must be in ascending order. The test passes for standard_user (where sorting works) and fails for problem_user (where it is broken) — turning a manual observation into an automated check that catches the defect on every run.

Tech stack

Playwright · pytest · pytest-playwright · GitHub Actions

Getting started

pip install -r requirements.txt
playwright install chromium
python -m pytest

Project structure

tests/test_sorting.py   # automated regression test (Playwright + pytest)
docs/bug-reports.md     # test plan + documented defects
requirements.txt        # dependencies
.github/workflows/      # CI (runs the tests on every push)

Notes

The automated tests run against the live SauceDemo site, so they depend on that site being reachable and unchanged. A failure can therefore reflect an external issue (site downtime or a change on Sauce Labs' side) rather than a problem in this repository — a trade-off worth keeping in mind for tests that target a third-party application.