Choose a meaningful combination
A keyboard may report one combination of held keys differently from another. Begin with the exact group you use in a game, shortcut or application rather than trying to press every key at once. Start the focused test and press the keys one by one while continuing to hold the earlier ones. The board identifies the codes currently observed as held.
Release the group deliberately and inspect the record. The peak simultaneous value is useful evidence of that sequence. It is not a certificate that every possible group of the same size works, and it cannot prove that a key the browser never received was physically pressed. Repeat the important combinations in a consistent order.
Account for shortcuts and layout differences
Some key combinations are reserved by the browser or operating system. They may move focus, change tabs, open a system interface or close a window. The page cannot guarantee capture of these combinations. Tab intentionally leaves the capture surface, and Escape ends a session so you always have a keyboard route out of the test.
The board uses physical-position code names. A name such as KeyQ is not necessarily the character produced on AZERTY, Dvorak or another layout. The display includes extra ISO and JIS positions as well as common ANSI codes. Keys absent from your particular keyboard are simply unused positions; the tool does not interpret their unlit state as a defect.
Keep focus interruptions visible
If focus leaves the area or the page becomes hidden, the session stops. Keys still held at that moment can remain listed as held at the end because their release was outside the recorded interval. That is an incomplete observation, not proof that the key is electrically stuck. Check the end reason alongside the event list.
For a comparison, begin another fresh session and reproduce the same chord while keeping the page in the foreground. Use the previous stopped session as a baseline or load its JSON file. The summary helps compare observed counts and candidate events, but different pressing routines and different session lengths remain different experiments. Keep the underlying sequence when the details matter.