About GearLint
GearLint is a free set of fault tests for mice, keyboards and controllers — find out whether a fault is really hardware, then decide whether to repair, replace or return the device. Everything runs in your browser: no downloads, no accounts, no data leaving your device.
Who builds GearLint
Why this site exists
Peripherals fail quietly and expensively. A drifting thumbstick gets blamed on game settings; a double-typing key gets blamed on the user. The web already ships the APIs needed to see the raw truth — Gamepad, keyboard and pointer events — but most test sites bury them under ads, installers and sign-up walls. GearLint takes the opposite approach: the test is the page, it runs on standard browser APIs, and when you close the tab your data is gone, because it never went anywhere.
The name is borrowed from software culture: a linter is a tool that scans code for faults automatically, before they cause damage. I do the same for your gear — lint it before you blame a game, a driver, or your wallet.
How the tests work
Every test reads a browser API and prints what it returns, without interpretation you cannot check yourself. The controller test shows raw axis values, so a resting stick reading 0.09 is visible as 0.09 rather than hidden behind a pass or fail badge. The mouse test measures the time between press and release events to expose the double-click bug, and counts events per second for a polling-rate estimate. The keyboard test compares the gap between a keydown and its keyup against a threshold you can see and change.
Two limits are worth stating plainly, because honest tools mention them. A browser is not an oscilloscope: pointer events are delivered on animation frames, so a 1000Hz mouse will read far lower here than on a dedicated polling tool, and that number is a responsiveness check rather than a specification check. And nothing here can see inside a device — a test can prove a switch bounces, not that a specific contact is worn.
How the guides are written
Each guide is built around what the matching test actually measures: which number decides the verdict, which threshold applies, and what to do with the result. Repair steps describe procedures that are widely documented and repeatable — cleaning a potentiometer, pulling a hot-swap switch, cycling a stuck pixel. Where an outcome depends on the exact model, I say so instead of inventing a universal answer, and where a fix is temporary I call it temporary rather than burying the caveat.
If a guide turns out to be wrong or out of date, tell me and it gets corrected — the contact page explains what to include.
What runs where
Every test executes in your browser tab using standard web APIs. There is no backend server receiving your keystrokes, axis values or click timings — see the privacy policy for the full breakdown. Advertising, once enabled, is served in designated slots by third-party networks under their own policies.
What this site does not do
Nothing here asks you to create an account, install anything, or trust a browser extension. I do not sell or share your data, and I do not accept paid placements inside guides — affiliate links, where they appear, are labelled and never change a recommendation.
Contact
Questions, bug reports, or partnership inquiries: hello@gearlint.com. For privacy-specific questions, use privacy@gearlint.com. More detail on both is on the contact page.
GearLint · Find the fault. Decide what to do about it.