diff --git a/plans/audio-review-fixes-handoff.md b/plans/audio-review-fixes-handoff.md
index 0015683e6c..3b553cffce 100644
--- a/plans/audio-review-fixes-handoff.md
+++ b/plans/audio-review-fixes-handoff.md
@@ -32,12 +32,32 @@ plain fast-forward.
 packages/studio    3662 passing, 18 todo, 1 file skipped
 packages/core      1669 passing
 packages/engine     736 passing, 3 skipped (src/services)
-packages/producer   534 passing — plus 50 PRE-EXISTING collect failures
+packages/producer   534 passing under vitest, 203 under bun test (see below)
 ```
 
-The producer collect failures are files importing `bun:test`, which `vitest run`
-cannot load. Same 50 before and after this work; only the passing count moved
-(533 → 534). Producer was never in the original baseline set, which is why.
+**Producer's suite is split across two runners.** 49 files import `bun:test`
+and `vitest run` cannot load them — it reports them as collect failures, which
+is a harness split, not breakage. Run BOTH:
+
+```bash
+cd packages/producer
+./node_modules/.bin/vitest run          # 534 passing, 50 files fail to collect
+bun test ./src/services/distributed/plan.test.ts ./src/plan-parity-*.test.ts \
+         ./src/services/render/stages/*.test.ts        # 203 passing
+```
+
+Note the leading `./` — without it `bun test` reads the arguments as name
+filters and runs nothing while exiting 0. A bare `bun test` is worse than
+useless here: it also picks up the vitest files, which it cannot run, so it
+reports ~37 phantom failures.
+
+The bun half matters for the `b1faacb3a` audioStage change specifically:
+`plan.test.ts` and the `plan-parity-*` suite cover
+`applyDistributedAudioWarningPolicy`, the consumer whose input that commit
+changed. Verified green.
+
+Producer was never in the original baseline set, which is why neither half was
+tracked before.
 
 `tsc`, `oxlint`, `oxfmt` clean across studio, core, engine, producer, lint.
 
@@ -125,6 +145,12 @@ Three judgement calls worth knowing:
   yet.** The numbers: the chain peaks at 1.346, the old order sheared 19,166
   samples (1.7% of the file) flat against ±1, and level-matched the two differ
   by an error signal 26.3 dB down with a worst single-sample delta of 0.242.
+  The generator sits beside them — `bun ~/audio-fx-clip-ab/ab-clip.mts <outdir>
+  <source.wav>` — so the pair can be rebuilt with a different chain, a
+  different duck, or the user's own audio. It drives the real
+  `applyAudioFxChain`, once with the envelope and once without followed by the
+  mixer's old second pass, so "before" is the actual previous behaviour rather
+  than a simulation of it.
 - **§4a-4 striding was tried and rejected.** Capping Welch at N windows is 27x
   faster on a 5-minute voiceover (427 ms → 15 ms) but *changes the carve*: at
   strength 0.9 the selected band set moved (630 Hz → 160 Hz) and it did not