R30 — CLAUDE.md's ~31% coverage figure predates 26 new JVM test files #39

Closed
opened 2026-08-23 03:44:54 +00:00 by JMR-dev · 1 comment
JMR-dev commented 2026-08-23 03:44:54 +00:00 (Migrated from github.com)

Finding R30 from the overnight max-effort review (Fable lead, Opus sub-agents). Full report: scratchpad/overnight/REVIEW.md.

R30 — CLAUDE.md's ~31% coverage figure predates 26 new JVM test files

severity: low
verdict: PLAUSIBLE (direction certain, magnitude unmeasured)
where: CLAUDE.md:99-100; echoed docs/defect-audit.md:53-54, :849-850
scenario: The figure is the argument against a coverage floor. Accurate when computed (jacoco XML 14:18:35 = 31.0%) but the suite then grew 180 -> 257 across exactly the ~1,200 untested lines the 31% was derived from.
evidence: XML date + line counts; jacoco not re-run (hence PLAUSIBLE).
fix: Re-run :app:jacocoTestReport, update all three sites together. Decide whether a settled higher number now argues FOR a floor.
risk: One Gradle run.


Cut: above — worked autonomously overnight.

🤖 Generated with Claude Code

_Finding **R30** from the overnight max-effort review (Fable lead, Opus sub-agents). Full report: `scratchpad/overnight/REVIEW.md`._ ### R30 — CLAUDE.md's ~31% coverage figure predates 26 new JVM test files severity: low verdict: PLAUSIBLE (direction certain, magnitude unmeasured) where: CLAUDE.md:99-100; echoed docs/defect-audit.md:53-54, :849-850 scenario: The figure is the argument against a coverage floor. Accurate when computed (jacoco XML 14:18:35 = 31.0%) but the suite then grew 180 -> 257 across exactly the ~1,200 untested lines the 31% was derived from. evidence: XML date + line counts; jacoco not re-run (hence PLAUSIBLE). fix: Re-run :app:jacocoTestReport, update all three sites together. Decide whether a settled higher number now argues FOR a floor. risk: One Gradle run. --- **Cut:** `above` — worked autonomously overnight. 🤖 Generated with [Claude Code](https://claude.com/claude-code)
JMR-dev commented 2026-08-23 05:51:26 +00:00 (Migrated from github.com)

Fixed in #50 (merged). Measured, not assumed: 29.8% lines (629/2113), 28.7% branches. The figure FELL from ~31% despite 11 -> 43 test files, because main source grew 4,114 -> 5,715 lines. This finding's stated direction ("certain") was backwards.

Fixed in #50 (merged). Measured, not assumed: 29.8% lines (629/2113), 28.7% branches. The figure FELL from ~31% despite 11 -> 43 test files, because main source grew 4,114 -> 5,715 lines. This finding's stated direction ("certain") was backwards.
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: JMR-dev/LibreMediaConverter#39