Sample report
Sample report. Real probe of the SauceDemo test store, not a customer.
Accessibility evidence report · sample
saucedemo.com
- Client
- Sample only
- Prepared
- September 2026
- Scope
- Public + signed-in
- Engine
- axe-core 4.13.0
Executive summary
The public pages scanned clean. 3 violations were waiting in states that only exist once you sign in, sort the catalog, or trip a validation error.
- Critical
- 3
- Serious
- 0
- Moderate
- 0
- Minor
- 0
Fix first
- 1Criticalselect-name
Blocks the sort control on the main product list, the first thing a signed-in shopper touches.
- 2Criticalbutton-name
Blocks recovery from a checkout error, the highest-cost place to lose a customer.
Findings · signed-in states
- Criticalselect-name
A screen reader can't tell a person what the dropdown selects.
Fix: Associate a visible <label> with the <select>, or add an aria-label naming what it controls.
Learn more about select-name on Deque University (opens in a new tab)
found at saucedemo.com · inventory (behind auth)
1 additional critical finding is logged in the raw probe below but not excerpted here.
Findings by state · public vs. behind-login
A URL-level scanner only ever sees the first row. The rest exist only after a session is established.
- 01Public
Login
No violations
- 02Session only
Login error
Critical
- 03Session only
Inventory
Critical
- 04Session only
Cart
No violations
- 05Session only
Checkout
No violations
- 06Session only
Checkout error
Critical
Raw results and method: experiments/net-new-violations
Persona layer
The report says what broke. This shows who it stopped.
Persona replay · saucedemo.com
· step 1 of 3“Bad credentials. Looking for what went wrong.”
Illustrative walk from experiments/net-new-violations. The task success opinion is separate from deterministic WCAG findings.
Method & limitations
- Findings come from axe-core 4.13.0 run at every state a signed-in crawl reaches: page load, after login, after adding to cart, and on a validation error. Deterministic: run it twice, get the same answer.
- Automated checks cover a large share of WCAG failures but not all of them. Manual keyboard and screen-reader review still matters; this report is not a substitute for testing with disabled users.
- Persona task-success is an AI browser agent's read of a final page state and its own written opinion, never a WCAG verdict, and never mixed into the findings above.
- This is a public sample built from a real, published probe. It is not a customer account and no credentials of any kind were used or stored.
Want this for your own site? Grade the public pages free, or see what an agency plan includes. Behind-login scanning runs in the CLI.
Target: saucedemo.com · Source: experiments/net-new-violations