Let a button press cut the current reply (barge-in)

Every press bumps a turn counter and kills the running aplay; queued
audio from an older turn is dropped, and reply() now runs on its own
thread and returns once its turn is stale, so the button loop is never
blocked. A cut-off exchange is left out of history so user and
assistant messages still alternate.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

1 files changed, 22 insertions, 2 deletions

diff --git a/AGENTS.md b/AGENTS.md
index bfc29f4..0e58c38 100644
--- a/AGENTS.md
+++ b/AGENTS.md
@@ -362,6 +362,28 @@ file: the model fills parameters but cannot reach `mailctl tag` or
   confirmation before acting for exactly this reason.
 - These tools speak personal data (mail subjects, amounts) out loud.
 
+### Barge-in (2026-09-26)
+
+A button press cuts the current reply. A global `turn` counter goes up on
+every press; queued audio items carry their turn and the speaker drops
+stale ones; the press kills the running `aplay` at once (that also drops
+the ~1 s already in its pipe). `reply()` runs on its own thread and
+returns when its turn is stale: after STT, between streamed pieces,
+before tool rounds. The button loop is never blocked by a reply any more.
+
+- Live: stopped instantly while speaking and while a search was running,
+  nothing of the old reply played afterwards (by ear). A correction
+  ("anzi, dimmi...") pressed during a search got its own clean answer.
+- A reply whose text had already finished generating is cut in audio
+  only; the log shows no `interrupted` then, and the exchange stays in
+  history. A reply cut while generating is dropped from history whole,
+  so user and assistant messages still alternate.
+- Known ceilings (`ponytail:` comments): a running tool is not killed,
+  its result is dropped; a press during the old reply's STT can
+  overwrite the WAV mid-upload.
+- Unrelated but seen: whisper garbled "i Mille" into "i 1000.000e" and
+  the model invented an answer for it instead of searching.
+
 ### TTS (2026-09-25)
 
 `piper-tts` on CPU (onnxruntime), ~5 s sentence per voice:
@@ -418,8 +440,6 @@ mic, TTS, glue); push-to-talk came for free from the mic button. Open, in
 no particular order:
 
 - **Run as a service/autostart** (it currently runs by hand).
-- **Barge-in:** pressing the button while it speaks does not stop the
-  reply yet.
 - **Check presentation mode during games:** `statusctl presentation get`
   reads the stored flag; confirm it is 1 while a game holds it on.
 - **Latency:** STT ~1.2 s is the biggest share; the double encoder pass