PRODUCTS & SYSTEMS · 2026 · PRIVATE TEST · ADOPTION UNVERIFIED
The most consequential product decision in Chejillog was a refusal. The app could hold observations that a person wanted to revisit, but it would not determine a constitutional type or offer a diagnosis or prescription.
Source code, an Android bundle, program-registration records and a private-test trail survive. I have not verified a current public store listing, real users, adoption or revenue. This is therefore a build note, not a launch story.
A smaller interface made the boundary clearer
I reduced the main path to a home view, a place to record an observation and a way to look back. The stored value remained the user’s own note; the display could make it easier to revisit without turning it into a health conclusion. The copy was revised alongside the interface so that the app did not sound more authoritative than it was.
The Android build exposed a mundane failure
A backup that downloaded in a browser did not save from the Android web view. I replaced the browser-dependent path with a flow that created the file locally and passed it to the operating system’s share function. The private build was tested through that change.
Registration did not establish use
The registration record supports the existence of the software artifact. It does not show that anyone installed it, returned to it or benefited from it. Older working notes also contain store-release language that has not been reconciled with a currently visible listing, so I do not describe Chejillog as publicly released.
A responsible next step would begin with a fresh privacy review using synthetic records, followed by a direct check of distribution and actual use. Until then, the defensible outcome is narrow: a registered program and a privately tested Android build.
Project record
Period: 2026
Role: scope, interface and backup-flow design; implementation; private testing
Activity: built and tested an Android self-observation app
Internal evidence: source, build artifacts and test records
External validation: program-registration records
Final outcome: public distribution, users, benefit and revenue remain unverified