Example retest: the same checkout after the fix

What the run did

• Target: /demo/shop-fixed/ on Sybqa’s own server, run locally on 127.0.0.1 for this recording. Trailhead Supply is not a real shop: nothing is sold, sent or charged.
• Browser: WebKit 27.2, at Desktop 1280 × 800 and Narrow 320 × 740.
• Plan: the deterministic planner, from the goal “Buy a product and check out”. AI review and AI routing were off, and the plan did not allow writes.
• Recorded: 2026-11-05 01:07:43 UTC, 281 checks and 9 run notes in total.

Journey: Buy a product and check out

1. click [data-testid="product-link-trail-mug"]
2. expectVisible [data-testid="product-title"]
3. click [data-testid="add-to-cart"]
4. expectText [data-testid="cart-count"] with “1”
5. click [data-testid="view-cart"]
6. expectVisible [data-testid="cart-line"]
7. click [data-testid="checkout-link"]
8. fill #checkout-name with “Demo Shopper”
9. fill #checkout-email with “shopper@example.com”
10. fill #checkout-card with “4242 4242 4242 4242”
11. click [data-testid="place-order"]
12. expectVisible [data-testid="order-confirmation"]

Journey: Checkout validation

1. click [data-testid="place-order"]
2. expectVisible [data-testid="checkout-errors"]

The typed name, email and card number are the demo store’s own published test values.

Journey results

Screenshots

What changed

The fixed copy of the store, /demo/shop-fixed/, serves the same files as the seeded copy, /demo/shop/, with the checkout bug fixed. Both runs used the same goal, planner, browser and screen sizes.

The seeded run recorded 271 checks and 8 run notes, and this retest 281 checks and 9 run notes, so the lists are not identical: 14 IDs appear only here and 3 only in the seeded run. The fixed checkout reaches the order confirmation, which adds that step’s own checks. A journey’s runtime check is recorded only when the browser logs an error: the seeded checkout logged the planted bug’s error and the fixed one logged none. Some checks are recorded only when there is something to report, such as actions that showed no visible change.

One difference from a real retest: the fixed store has its own path so both runs can be published side by side. On your own site, rerun the same check at the same URL, screen size and starting state, and keep the failed run in the history.

What this run did not cover

• Not run (10 checks): Some page timing metrics are unavailable in this browser.
• Not run (2 checks): slow-network: network throttling needs Chromium; this run uses webkit.
• Not run (2 checks): double-submit: double-clicking "Add to cart" sent no submission request; nothing to deduplicate.
• Run notes, not counted as Not run checks:
  • Additional AI verification was not enabled for this check.
  • LLM exploration disabled by profile.
• No payment, order or email exists: the demo store has no back end, so these checks cover what the browser showed, not stored data.
• One fictional store, one browser engine and two screen sizes. It says nothing about any other site and is not a measure of how often Sybqa finds bugs.
• This recording used WebKit at 1280 and 320 px. Runs started from the console use Chromium at 1440, 390 and 320 px, so browser-specific checks such as slow-network throttling and page timing metrics can differ.

Every recorded check and run note

All 281 checks and 9 run notes from this run, each with the state and summary it recorded, word for word.

Run it yourself

The console can plan a check of the same demo store, or of a site you own or may test. Add your link here; it waits in the console after you create an account. The free plan gives 10 runs a month without AI review. Nothing runs until you review the plan.

Related

• Example run: a planted checkout bug on a demo store
• Example runs on a fictional demo store
• How to read a website QA report
• How to smoke-test a published website and retest a fix
• Bug report template: steps, evidence and retest status
• How to retest a website fix without writing test code
• Blocked is not passed: how to read open checks in a QA report
• Sample report with a recorded signup fixture