Test a size guide, a sold-out size and the size carried to the bag

The flow under test

A size selection is only trustworthy when the chart, the dialog, the sold-out state and the bag all agree, and each of those fails differently.

• The conversion table is correct - US 9 maps to UK 8, EU 42.5 and CM 27 on the same row.
• Selecting from the guide works - choosing a size inside the dialog updates the product page behind it.
• Sold out is enforced - US 9.5 is shown as disabled rather than selectable.
• The size reaches the bag - the bag line reads Black / Volt with Size US 9, colourway and size together.

Asserting a full row of the conversion table is unusual and worth doing. A size chart with one wrong European conversion sends the wrong shoe to every customer who trusts it, and the page gives no sign anything is off.

The test file

The size is selected from inside the dialog rather than from the product page grid, which is the path that is more likely to break and less likely to be tested by hand.

---
mode: testing
url: https://my-testing-repo-main.vercel.app/nike-clone-app?reset=true
max_steps: 45
tags: [nike, e-commerce, sizing]
---

# Nyke 8.1: Size guide and selection

Catalog objective: open the size guide on a shoe, select a size, and add it to the bag.
Key assertion: the selected size is carried into the bag.

## Open a running shoe
Click "Nyke Aeroglide 41" and verify the product page shows the colourway "Black / Volt" and the price "$139.99".

## Open the size guide
Click "📏 Size guide" and verify a dialog appears titled "Size guide" showing a row where US "9" maps to UK "8", EU "42.5" and CM "27".

## Select a size from the guide
Click "Select" on the US "9" row and verify the product page shows a badge reading "Selected: US 9".

## Confirm a sold-out size cannot be chosen
Verify the size "US 9.5 ✕" option is shown as disabled.

## Add to bag and confirm the size carried
Click "Add to Bag", click the "Bag" link in the header, and verify the bag shows "Nyke Aeroglide 41" with "Black / Volt · Size US 9".

The final step asserts colourway and size in a single string, so a bag that keeps the size and loses the colourway fails. Run the test with:

kane-cli testmd run tests/nike/size-guide_test.md

The run

Five passes. Checking the sold-out size was the cheapest step in this entire set at 18.1 seconds, since it is a single state assertion on a page already loaded.

---
test: ../size-guide_test.md
status: passed
started: 2026-09-15T10:39:14.068Z
duration_s: 190
session_id: 7e3db78a-761c-4617-9fde-c264d7fc13dc
---

# Nyke 8.1: Size guide and selection - Result

## Open a running shoe ✓ passed (55.4s)
md5: e17fd5491c1a286889a061534d888082
Click "Nyke Aeroglide 41" and verify the product page shows the colourway "Black / Volt" and the price "$139.99".

## Open the size guide ✓ passed (25.7s)
md5: 9022f1e74f93e29ca92a8f4e9fde15a9
Click "📏 Size guide" and verify a dialog appears titled "Size guide" showing a row where US "9" maps to UK "8", EU "42.5" and CM "27".

## Select a size from the guide ✓ passed (34.7s)
md5: 37c37ce5e68f9443ee9f2f990577edfb
Click "Select" on the US "9" row and verify the product page shows a badge reading "Selected: US 9".

## Confirm a sold-out size cannot be chosen ✓ passed (18.1s)
md5: a74330e8d69635aeaa500f2794a1f6e7
Verify the size "US 9.5 ✕" option is shown as disabled.

## Add to bag and confirm the size carried ✓ passed (53.8s)
md5: 4a669118bd7bb7ec611e4d326f68a91a
Click "Add to Bag", click the "Bag" link in the header, and verify the bag shows "Nyke Aeroglide 41" with "Black / Volt · Size US 9".

An 18 second assertion is a reminder that negative checks are nearly free once the page is open. There is rarely a good reason to leave them out.

Reading the evidence

The result is sizing coverage that includes the conversion data itself, which is the part that causes returns and the part no manual pass ever re-reads.

Add one row assertion per size you sell most of, and the chart is covered. Generating the variants makes that cheap, and the pre-PR check is where a size grid change should be caught.