CLAUDE.md says re-measure before quoting, and its own figure was measured on 2026-08-24 —
before the #52 children, the MediaProbe/codec tests, and the #100/#107 guards landed.
Measured on b53f326 with ./gradlew :app:jacocoTestReport:
counter
now
was (2026-08-24)
LINE
1847/2268 — 81.4%
1519/2194 — 69.2%
BRANCH
60.2%
53.2%
417 JVM tests in 60 classes, all green.
The denominator moved too, 2194 → 2268, because the same push added production code of its own.
The note now says that rather than presenting this as a pure numerator gain.
The Robolectric / isIncludeNoLocationClasses history is left exactly as it was — it explains why
every pre-2026-08-24 figure was an artifact, and it is still the most useful thing in the entry.
The phrase "before that date" is now spelled out as "before 2026-08-24", since the headline date
above it has moved and the phrase no longer points at itself.
Docs only; no behaviour change.
`CLAUDE.md` says **re-measure before quoting**, and its own figure was measured on 2026-08-24 —
before the #52 children, the MediaProbe/codec tests, and the #100/#107 guards landed.
Measured on `b53f326` with `./gradlew :app:jacocoTestReport`:
| counter | now | was (2026-08-24) |
|---|---|---|
| LINE | **1847/2268 — 81.4%** | 1519/2194 — 69.2% |
| BRANCH | **60.2%** | 53.2% |
417 JVM tests in 60 classes, all green.
**The denominator moved too**, 2194 → 2268, because the same push added production code of its own.
The note now says that rather than presenting this as a pure numerator gain.
The Robolectric / `isIncludeNoLocationClasses` history is left exactly as it was — it explains why
every pre-2026-08-24 figure was an artifact, and it is still the most useful thing in the entry.
The phrase "before that date" is now spelled out as "before 2026-08-24", since the headline date
above it has moved and the phrase no longer points at itself.
Docs only; no behaviour change.
Held deliberately, not stalled. Auto-merge disarmed.
The figure in this PR was measured on b53f326. #113 is armed and adds production code plus ~300
lines of tests, and #30 is still in flight. Merging now would commit a number that is stale within
the hour — which is precisely the failure this PR exists to correct, and CLAUDE.md's own rule is re-measure before quoting.
I will re-run :app:jacocoTestReport once both have landed, push the settled figure, and arm it
then. The measurement commit is named in the body so whatever lands can be checked against it.
**Held deliberately, not stalled.** Auto-merge disarmed.
The figure in this PR was measured on `b53f326`. #113 is armed and adds production code plus ~300
lines of tests, and #30 is still in flight. Merging now would commit a number that is stale within
the hour — which is precisely the failure this PR exists to correct, and `CLAUDE.md`'s own rule is
**re-measure before quoting**.
I will re-run `:app:jacocoTestReport` once both have landed, push the settled figure, and arm it
then. The measurement commit is named in the body so whatever lands can be checked against it.
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
CLAUDE.mdsays re-measure before quoting, and its own figure was measured on 2026-08-24 —before the #52 children, the MediaProbe/codec tests, and the #100/#107 guards landed.
Measured on
b53f326with./gradlew :app:jacocoTestReport:417 JVM tests in 60 classes, all green.
The denominator moved too, 2194 → 2268, because the same push added production code of its own.
The note now says that rather than presenting this as a pure numerator gain.
The Robolectric /
isIncludeNoLocationClasseshistory is left exactly as it was — it explains whyevery pre-2026-08-24 figure was an artifact, and it is still the most useful thing in the entry.
The phrase "before that date" is now spelled out as "before 2026-08-24", since the headline date
above it has moved and the phrase no longer points at itself.
Docs only; no behaviour change.
Held deliberately, not stalled. Auto-merge disarmed.
The figure in this PR was measured on
b53f326. #113 is armed and adds production code plus ~300lines of tests, and #30 is still in flight. Merging now would commit a number that is stale within
the hour — which is precisely the failure this PR exists to correct, and
CLAUDE.md's own rule isre-measure before quoting.
I will re-run
:app:jacocoTestReportonce both have landed, push the settled figure, and arm itthen. The measurement commit is named in the body so whatever lands can be checked against it.