From bca886c84441fae08a0135654b246e7f57d2a737 Mon Sep 17 00:00:00 2001
From: kshitijk4poor <82637225+kshitijk4poor@users.noreply.github.com>
Date: Sun, 19 Jul 2026 07:21:43 +0530
Subject: [PATCH] fix(gateway): invalidate pending STT cache when media merges
 into event

merge_pending_message_event extends the existing event's media_urls in
place when two media-bearing messages arrive in quick succession (photo
bursts, consecutive voice messages).  The gateway runner caches STT
transcripts on the event via _gateway_pending_stt_text; if the cached
event gains new media after the cache was populated, the stale transcript
was returned instead of transcribing the merged attachments.

Add _invalidate_pending_stt_cache() and call it from both media-merge
branches in merge_pending_message_event so the next transcription call
re-runs against the full merged media list.

Closes the merge-race edge case identified during review of PR #61519.
---
 gateway/platforms/base.py | 21 +++++++++++++++++++++
 1 file changed, 21 insertions(+)

diff --git a/gateway/platforms/base.py b/gateway/platforms/base.py
index 1391395c32ff..4abe92072c68 100644
--- a/gateway/platforms/base.py
+++ b/gateway/platforms/base.py
@@ -2104,6 +2104,25 @@ class EphemeralReply(str):
         return str.__str__(self)
 
 
+def _invalidate_pending_stt_cache(event: MessageEvent) -> None:
+    """Clear gateway-side STT cache attrs when media is merged into an event.
+
+    ``merge_pending_message_event`` extends ``media_urls`` in place when two
+    media-bearing messages arrive in quick succession.  The gateway runner
+    caches STT transcripts on the event via ``setattr`` (see
+    ``_transcribe_pending_audio_event_once``); if the cached event gains new
+    media after the cache was populated, the stale transcript must be
+    discarded so the next transcription call picks up the merged attachments.
+    """
+    for attr in (
+        "_gateway_pending_stt_text",
+        "_gateway_pending_stt_transcripts",
+        "_gateway_pending_stt_echo_sent",
+    ):
+        if hasattr(event, attr):
+            delattr(event, attr)
+
+
 def merge_pending_message_event(
     pending_messages: Dict[str, MessageEvent],
     session_key: str,
@@ -2134,6 +2153,7 @@ def merge_pending_message_event(
             existing.media_types.extend(event.media_types)
             if event.text:
                 existing.text = BasePlatformAdapter._merge_caption(existing.text, event.text)
+            _invalidate_pending_stt_cache(existing)
             return
 
         if existing_has_media or incoming_has_media:
@@ -2152,6 +2172,7 @@ def merge_pending_message_event(
                 and event.message_type != MessageType.TEXT
             ):
                 existing.message_type = event.message_type
+            _invalidate_pending_stt_cache(existing)
             return
 
         if (