ACCESSIBILITY REGRESSION TESTING

See what a release sounds like before users do.

AccessTrace compares real HTML states and finds keyboard and screen-reader regressions that visual snapshots miss.

access-trace / checkout

✓ 14 keyboard stops traced

✓ 4 landmarks mapped

! 2 regressions found

Previously“Place order, button”Now“Button, unlabeled”
01 · LOCAL PLAYGROUND

Find the regression, not just the rule.

Choose a scenario or edit the candidate markup. Everything runs locally in your browser.

No data leaves this device
BASELINE · accessible release
CANDIDATE · current branch

COMPARISON RESULT

No new regressions found

Safe to merge

KEYBOARD JOURNEY

0 0focusable stops

    SCREEN-READER REGRESSIONS

    The candidate preserves the baseline accessibility journey.

    SEMANTIC MAP

    0 landmarks 0 headings 0 total findings

    What a screen reader hears first:

    No focusable element found.
    02 · HOW IT WORKS
    01

    Trace the keyboard journey

    Maps every reachable control in the same order a keyboard user encounters it.

    02

    Build the screen-reader journey

    Reads roles, names, headings, landmarks, labels, and dialog semantics.

    03

    Compare change, not noise

    Surfaces only new accessibility failures introduced by the candidate.