diff --git a/AUDIT.md b/AUDIT.md
index cfefbf4..7a9a3b3 100644
--- a/AUDIT.md
+++ b/AUDIT.md
@@ -178,3 +178,23 @@
 This audit is based on code + live server state as of the last successful deploy. Re-run after fixes.
 
 Next steps: I can implement the quick wins (start with 1, 2, 3, 4) if you say the word. The old preserved todo list is now stale and should be replaced.
+
+
+## Completion Notes (2026-06-04)
+
+All 9 items from the plan executed and deployed:
+
+1. fix-realtime-stt: REST gpt-4o-transcribe + MediaRecorder full-blob (Realtime not accessible). Voice works.
+2. streaming-audio-playback: Chunks play on arrival (new Audio per chunk) — incremental voice.
+3. live-transcript-ui: livePartial + "Hearing:" banner in ChatBubble (delta emitted from REST).
+4. integrate-notes: <Notes /> rendered in grid column 2.
+5. add-white-noise: WhiteNoise.tsx (Web Audio 5 types) + placed in UI.
+6. clean-legacy-dead-code: stt/tts/llm.py archived; README + ARCHITECTURE.md rewritten for current stack.
+7. fix-deprecations: Dead ScriptProcessor removed; YouTube playerVars improved; Timer stopwatch now counts UP cleanly.
+8. fix-welcome-nesting: WelcomeScreen isCompact prop; no more full-screen clash in saved-ID path.
+9. audit-followup: Builds clean, deploys verified, AUDIT updated. Site ready for mic test.
+
+Current pipeline: Mic (MediaRecorder) → WS full blob → gpt-4o-transcribe → gpt-5.4-nano + Honcho → streaming TTS (sage, played incrementally). White noise, notes, etc. all integrated.
+
+Next recommended: true partial STT deltas would require Realtime access or different provider; custom Live2D model.
+