Most tools stop at the web and hand you a manual checklist for your app. Wholisphere drives a real Android device through its own accessibility tree, pixels and gestures — and returns an automated result for every one of the 55 WCAG 2.2 Level A + AA criteria: a measured verdict, a media evaluation, or a documented Not Applicable. None are punted to a manual attestation. In-house evaluators, no third-party engine, and it never fakes a pass.
Developer preview Android, CLI-driven against your build. iOS and hosted scanning are on the roadmap.
A native app exposes a sparse accessibility tree — no heading roles, no autofill hints, no live-region markers. So we test the way a person and a screen reader actually experience it: real focus traversal, real gestures, real device state — not markup parsing.
Most mobile "audits" mark a dozen-plus criteria "not evaluated" and mail you a checklist. We turned each of those into a probe, a measurement, or a judged verdict — so your VPAT reports a disposition for every criterion, not a to-do list.
The scanner drives a real device and reads it the way a tester does — through the accessibility tree, the pixels on screen, the device’s own gestures and sensors, and, where a rule can’t decide, a judge.
Deterministic checks straight off the platform’s own accessibility tree — names, roles, states, target size (≥24dp, density-aware), label-in-name, meaningful sequence — plus Otsu-measured text and non-text contrast from the screenshot. The exact surface a screen reader relies on, verified the way the OS exposes it.
Interaction probes drive a real device or emulator: TAB/switch traversal (keyboard reach, focus order, traps), font-scale and rotation re-capture (resize, reflow, orientation), a press-and-drag-off gesture (pointer cancellation), a two-capture countdown read (timing adjustable), on-focus/on-input context checks, and — on an emulator — injected accelerometer motion (motion actuation). “Cannot verify” becomes a measured verdict.
Vision-LLM rules on use-of-color, images-of-text, sensory instructions, and heading descriptiveness against the rendered screen. A per-screen frame stack screens for flashing (2.3.x) and auto-updating motion (2.2.2). A multimodal judge reads captions against the audio for the 1.2.x media criteria, and Unicode script analysis flags foreign-language passages. Measured, never assumed — no key means “unverified,” never a false pass.
Coverage, stated honestly: all 55 WCAG 2.2 Level A + AA criteria are dispositioned on native Android — 44 emit an automated per-screen verdict, 6 are media-evaluated when the app ships time-based media, and 5 are genuinely Not Applicable to a touch-native screen. None are left to a manual checklist. When a single scan honestly can’t be certain — a caption’s accuracy, a legal-flow safeguard — the finding is confidence-scored and routed to a one-click reviewer step, never rounded up to a false pass. The criteria map to the WCAG 2.2, Section 508 and EN 301 549 columns of the VPAT. Android today; the evaluator core is shared with the web scanner and shaped for iOS next.
Native mobile scanning is in developer preview — Android now, iOS next. Ask us for an invite.
