diff --git a/.claude/skills/preflight/SKILL.md b/.claude/skills/preflight/SKILL.md
index de1ca7f..d6ab83c 100644
--- a/.claude/skills/preflight/SKILL.md
+++ b/.claude/skills/preflight/SKILL.md
@@ -48,7 +48,7 @@ python3 .claude/skills/preflight/api37_e2e.py   # api37 preview E2E (hand-provis
 
 `jacocoTestCoverageVerification` runs right after `testDebugUnitTest` because it reads that
 task's JVM exec data — it enforces the whole-app **no-regression line-coverage floor (currently
-0.79)**, so a coverage regression is caught locally instead of only in CI (the exact class of
+0.84)**, so a coverage regression is caught locally instead of only in CI (the exact class of
 failure that reached CI on #367).
 
 `compileDebugAndroidTestKotlin` compiles the `androidTest` source set — the E2E/instrumented
diff --git a/CLAUDE.md b/CLAUDE.md
index ebf1eb2..da5a944 100644
--- a/CLAUDE.md
+++ b/CLAUDE.md
@@ -33,7 +33,7 @@ you changed via `python3 .claude/skills/preflight/local_instrumented.py <classes
 preview E2E via `python3 .claude/skills/preflight/api37_e2e.py` (the `/preflight` skill does all
 of this). `jacocoTestCoverageVerification` runs right after `testDebugUnitTest` (it reads that
 task's JVM exec data) and enforces the whole-app **no-regression line-coverage floor (currently
-0.79)**, catching coverage regressions locally instead of only in CI (the exact class of failure
+0.84)**, catching coverage regressions locally instead of only in CI (the exact class of failure
 that reached CI on #367). `compileDebugAndroidTestKotlin` compiles the `androidTest` source set
 that the static part of the gate skips, catching E2E/instrumented-test compile errors before
 they surface only in CI. `ktlintCheck`/`detekt` cover the `test`/`androidTest` source sets that
diff --git a/app/build.gradle.kts b/app/build.gradle.kts
index 416af1f..0e067b9 100644
--- a/app/build.gradle.kts
+++ b/app/build.gradle.kts
@@ -407,11 +407,17 @@ tasks.register<JacocoReport>("jacocoTestReport") {
 // No-regression coverage gate (closes #251; scoping from #290/#292). Fails `check` / CI when the
 // overall LINE coverage of the scoped surface above drops below `jacocoLineCoverageFloor`. This is a
 // FLOOR, not an absolute 95% target — the maintainer chose a ratchet over a fixed goal. The floor is
-// set a hair (~0.5–1%) below the measured baseline so ordinary run-to-run noise doesn't red-flag it,
-// while a real regression still fails the build. Manual ratchet FOR NOW: when coverage rises
-// materially, bump this number up in the SAME PR so the floor tracks reality (there is no auto-ratchet
-// yet). Baseline measured when this floor was set: 80.21% line (4838/6032), floor 0.79 (~1.2% headroom).
-val jacocoLineCoverageFloor = "0.79"
+// normally set a hair (~0.5–1%) below the measured baseline so ordinary run-to-run noise doesn't
+// red-flag it, while a real regression still fails the build. Manual ratchet FOR NOW: when coverage
+// rises materially, bump this number up in the SAME PR so the floor tracks reality (there is no
+// auto-ratchet yet).
+//
+// Re-ratcheted for #386 (final step of the Robolectric Compose epic #373, once infra/PoC #375 and
+// conversion batches #376-384 had all landed and proven stable): new baseline 87.89% line
+// (7994/9095), floor 0.84 — a wider ~3.9% headroom than the usual ~0.5-1%, chosen deliberately
+// conservative for this first post-epic measurement; the maintainer can tighten it further in a
+// follow-up PR. Prior baseline: 80.21% line (4838/6032), floor 0.79 (~1.2% headroom).
+val jacocoLineCoverageFloor = "0.84"
 tasks.register<JacocoCoverageVerification>("jacocoTestCoverageVerification") {
     // Same inputs as jacocoTestReport (shared vals above) so the gate enforces exactly what the
     // report shows. Depend on the unit tests so the exec data exists before verifying.