Our goal is to reduce the stress of uncertainty.
Before CheckMyARM can tell you anything useful, it has to look. The scanner reads what's actually installed and connected on your machine — not what you remember installing, not what a spec sheet claims, what's really there right now. Everything else follows from that starting point.
No single fact about an application tells us much on its own. An app being installed doesn't mean it matters to you. An app you haven't opened in months might still be essential once a year. So instead of asking one question and trusting the answer, CheckMyARM looks at several independent signals and lets them reinforce — or contradict — each other.
None of these, by itself, is proof that something matters to you. Together, they start to paint a picture — and the more of them that agree, the more confident that picture becomes.
This is a deliberate rule, not a technical detail: CheckMyARM never lets one piece of evidence single-handedly determine a recommendation. A single signal can be misleading — a device could be attached today by coincidence, an app could be open because you happened to click it a minute ago. It's the pattern across multiple independent observations that CheckMyARM actually trusts.
Automated observation only gets you so far. Some questions — does this specific app actually work well on Windows on ARM, not just launch — are best answered by someone who's already running it. CheckMyARM combines what it can observe on your machine with what real people have found on theirs, so the recommendation reflects more than one source of truth. See the About page for how that community actually works.
CheckMyARM can observe your machine, but it can't read your mind. It doesn't know that the app it flagged as low-priority is actually the one thing your job depends on, or that the one it assumed mattered is something you stopped using years ago and just never uninstalled. That's why the Workspace exists — a place to add what only you know, so your recommendation reflects your reality, not just what a scan can see.
A high-confidence recommendation doesn't mean "this is good news." It means CheckMyARM has enough independent evidence to trust its own reading of your situation — favorable or not. A low-confidence result isn't a failure either; it's an honest signal that there's real uncertainty left, and that a closer look (or a bit more of your own input) would genuinely help. Confidence is about how much we know, never a verdict dressed up as one.
What CheckMyARM looks at is scoped tightly to one question: is this software or device part of your computing environment? That boundary was drawn before anything was built, not added afterward to answer a complaint. See the Privacy page for exactly where that line is drawn.
Our goal is to reduce the stress of uncertainty.
Everything above is in service of that one idea — not a features list, just the reasoning behind
it.