Getyear proof standard · Current register
Proof should match the claim.
A passing build is not a shipped product. A simulator is not a physical device. A local archive is not an App Store release. We name each layer so you can see what is proven, what remains external and what decision the evidence supports.
This register reports the evidence stated by each case study. It is not a certification, independent audit or substitute for current production and real-user validation.- 04
- deep case registers
- 06
- distinct evidence layers
- CLEAR
- remaining external gates
01 / TapWar
A release system whose proof stops before public-market claims.
The evidence supports a working game and release presentation path; it does not turn implementation proof into adoption proof.
- Evidence
- 21/21 focused tests passed; five final store-presentation visuals exist; Release UI was exercised against production question and aggregate-result data.
- Boundary
- No public App Store availability, customer, revenue, retention or market-demand claim.
02 / BodySignal
Personal evidence with fixtures, device checks and clinical claims kept separate.
The proof covers a deterministic product foundation and one critical physical Apple Health comparison flow—not universal device readiness or health effectiveness.
- Evidence
- 56/56 canonical sleep-product tests passed; the critical physical Apple Health comparison flow was exercised; synthetic fixtures are visibly labelled.
- Boundary
- No clinical-effectiveness, complete device-coverage, App Store availability or real-user outcome claim.
03 / YEAR
A native archive proves a build—not distribution.
On-device OCR, speech indexing and local search were carried through build and device evidence while external distribution remained an explicit separate gate.
- Evidence
- XCTest, Debug and Release builds passed; a connected-device install and local Release archive were produced; demo media remains safe and labelled.
- Boundary
- No upload, TestFlight, App Review, public distribution, retention or revenue claim.
04 / Style intelligence
A paid-product path without pretending payment or model quality is live.
The case establishes an inspectable native report experience and implementation baseline; commerce and real-world quality remain outside that proof.
- Evidence
- 155 backend tests passed; Debug and Release simulator builds passed; sample report content is explicitly labelled.
- Boundary
- No live-payment, production-model-quality, App Store, conversion, retention or revenue claim.
05 / Evidence labels
What each proof layer actually means.
The useful question is not whether something looks finished. It is whether the evidence was collected at the layer where the important failure can occur.
- Tests
- Deterministic assertions passed in the stated suite. They prove only the behavior and cases that suite actually covers.
- Simulator
- The app built and ran in a simulated environment. Hardware, live permissions and external services may still differ.
- Device
- A named journey ran on physical hardware. One flow or device never implies complete coverage.
- Archive
- A local archive or packaging artifact was produced in the stated environment. Signing, export, upload, TestFlight, review and distribution remain separate unless individually verified.
- Production
- A real hosted system or production-shaped dependency answered the checked path at that time. Availability can drift.
- Public / user
- Only current distribution, commerce and representative real-user evidence can support market, conversion, retention or revenue claims.
06 / Non-inference
What we never infer from a green check.
Honest boundaries make product decisions faster because the next unknown is visible instead of hidden behind a broad readiness label.
- Simulator ≠ physical hardware or live permissions
- Local StoreKit ≠ App Store Connect or sandbox commerce
- Archive ≠ upload, TestFlight or App Review
- Passing tests ≠ customer demand, retention or revenue