Observed desktop inputs

InputActionEvidence scope
W A S DMovementArchived map tests recorded movement.
Mouse movementTurn / aimView changes were recorded.
Left clickFireBefore/after ammunition was inspected on ten maps.
Right clickAim down sightsInput exercised in desktop checks; weapon behavior can vary.
RReloadAmmunition replenishment was visible on seven of ten controlled map tests; do not infer completion from the keypress alone.
EscRelease pointer / pause interfaceUsed to leave captured input.
SpaceSkip an opening cinematicUsed during archived startup tests; this is not a verified general jump binding.

Make mouse capture work

  1. Use a desktop mouse or trackpad plus a keyboard. Touch-only play is not supported by this build.
  2. When gameplay is ready, click inside the canvas. The browser may require a direct click before it can capture the pointer.
  3. Press Esc to release the pointer before switching tabs or using browser controls.
  4. If the view jumps, adjust sensitivity in Options and retest in the same location. Change one setting at a time.

Interaction, sprint, crouch, jump and grenades

The October 18, 2026 browser checks observed F on purchase/power prompts and used it to collect Moon’s P.E.S.; X activated its helmet. In a later Moon check, F exchanged equipment for the Hacker and X held it for a successful 200-point door hack. These are separate actions: collecting equipment does not mean it is equipped. E was bound to lean right in the inspected configuration. Always follow your current on-screen prompt if bindings change.

Before following any objective guide, identify the current Interact action and test it on a harmless prompt. Do not assume that “press F” is correct until your build shows it. Rebinding support and persistence across reloads have not been verified.

Fullscreen and sensitivity

The original interface provides Options for sensitivity, volume and fullscreen. Browser fullscreen and pointer capture are separate permissions. Exit fullscreen with the browser’s usual control if you need to recover the page.

A fullscreen display can change the render resolution and workload. Record that change when comparing performance; a different viewport is a different test.

Sources and scope