Most screen-masking tools make the mask itself the thing you touch. You drop a black rectangle on your desktop, then you drag its edges, grab its corners, and nudge it into place by clicking the very object that is supposed to be covering something. It works until it doesn't. The mask sits on top of your windows, so now every click near it is ambiguous: did you mean to move the mask, or click the thing behind it? Tools paper over this with modifier keys, lock buttons, and "edit mode" toggles. That is the false-interaction trap. The cover and the controls live in the same surface, and they fight each other.
Screen Mask splits them apart.
The overlays that actually cover your desktop do one job: they cover. They are not draggable. They are not resizable by clicking them. They have no handles, no context menu, no hidden hotkeys. If you set one to click-through, your mouse passes straight to whatever is underneath as if the mask were glass. The overlay never competes with your real work for a click.
All the shaping happens somewhere else, in a control window that is a scaled map of your whole desktop. It is just a white rectangle with pixel rulers. Drag on it to draw a mask. Drag a drawn mask to move it. Pull its handles to resize. Right-click to delete. Because you are working on the map and not on the territory, there is never any doubt about what a click means. The map is for editing; the desktop is for covering. Two surfaces, two jobs.
This buys a few things almost for free:
- No accidental interaction. You cannot fat-finger a mask on your real screen, because the mask is not something you interact with there.
- Everything is visible at once. The control window lists every mask with its position, size, opacity, and click-through state. You are never hunting for a mask you lost behind another window.
- The desktop stays honest. Multi-monitor setups map cleanly, negative coordinates and all, because the preview represents the true virtual-screen geometry rather than a single display.
The tradeoff is that you look at a preview instead of dragging the real thing, and for the first ten seconds that feels indirect. Then it stops feeling like anything, the same way you stop noticing that a mouse moves a cursor and not your hand. What you get back is a tool that never argues with you about whether a click was meant for the mask or for the work behind it.
That is the whole idea. Keep the cover dumb. Put the intelligence in a place that can't get in your way.