fix: track escalation for legacy AI voice handoff path

request_handoff() (called by ai_voice_runtime_service for every real
call handoff, the only handoff path production calls actually use)
reserved an agent from the same routing pool as create_escalation()
but never created an EscalationRow, so the Phase 2 no-answer-retry
listener (DialEnd/Hangup) could never find it. A failed transfer
(no SIP registration, no answer, redirect error) left the agent
stuck in RESERVED forever with no retry to the next agent.

Now creates an EscalationRow (status=ringing) alongside the agent
reservation, releases the agent + marks the escalation failed if the
AMI Redirect itself errors immediately, and lets the existing
DialEnd/Hangup handler drive no-answer retry / release exactly like
the /escalations endpoint already does.

Reproduced live: call handed off to extension 2002 with no SIP
contact registered -> immediate hangup, cause=3, both pool agents
stuck in RESERVED indefinitely (had to release manually via psql).

deployment/aimaq.env.production

@@ -84,9 +84,9 @@ AI_VOICE_TTS_CACHE_DIR=/app/.data_local/ai_voice_tts_cache
 AI_VOICE_TTS_ELEVENLABS_API_KEY=sk_1c50faab05df13fd0ecde9aeca4abadcc7269ffdc15a03ad
 AI_VOICE_TTS_ELEVENLABS_API_BASE=https://api.elevenlabs.io
 AI_VOICE_TTS_ELEVENLABS_MODEL_ID=eleven_turbo_v2_5
-AI_VOICE_TTS_ELEVENLABS_RU_VOICE_ID=0ArNnoIAWKlT4WweaVMY
+AI_VOICE_TTS_ELEVENLABS_RU_VOICE_ID=ut2XM2wJyIZLTtW6lFzZ
 AI_VOICE_TTS_ELEVENLABS_RU_LANGUAGE_CODE=ru
-AI_VOICE_TTS_ELEVENLABS_KK_VOICE_ID=0ArNnoIAWKlT4WweaVMY
+AI_VOICE_TTS_ELEVENLABS_KK_VOICE_ID=ut2XM2wJyIZLTtW6lFzZ
 AI_VOICE_TTS_ELEVENLABS_KK_LANGUAGE_CODE=kk
 AI_VOICE_TTS_ELEVENLABS_OUTPUT_FORMAT=pcm_16000
 AI_VOICE_TTS_YANDEX_API_KEY=AQWJUMiUaXmbegxN4kgvM2XIlNqAPoBR5Wtq-40

services/ai_voice_runtime_service/media_runtime.py

@@ -482,7 +482,7 @@ class AudioSocketMediaRuntime:
         if ack_kind == "handoff":
             return (
                 "Угу, секунду.",
-                "Понял вас, соединяю.",
+                "Понялa вас, соединяю.",
                 "Мхм, соединяю.",
                 "Хорошо, сейчас соединю.",
                 "Да, сейчас соединю.",
@@ -492,7 +492,7 @@ class AudioSocketMediaRuntime:
             return (
                 "Угу, сейчас подскажу.",
                 "Мхм, сориентирую.",
-                "Ага, понял вас.",
+                "Ага, понялa вас.",
                 "Хм, сейчас уточню.",
                 "Хорошо, сейчас подскажу.",
                 "Ясно, сейчас разберусь.",
@@ -509,14 +509,12 @@ class AudioSocketMediaRuntime:
                 "Дайте уточню.",
             )
         return (
-            "Ага.",
             "Ясно, минутку.",
-            "Угу.",
             "Хорошо, сейчас.",
-            "Так, слушаю.",
-            "Момент.",
-            "Понял вас.",
-            "Хм, сейчас гляну.",
+            "Так, слушаю вас.",
+            "Один момент.",
+            "Поняла вас.",
+            "Хмм, сейчас гляну.",
         )
 
     def _select_ack_payload(

services/asterisk_bridge_service/voice_ai.py

@@ -311,7 +311,7 @@ def _resolve_handoff_extension(
     target_level: str | None = None,
     tenant_id: str | None = None,
     required_skills: list[str] | None = None,
-) -> tuple[str, str, str | None, str | None]:
+) -> tuple[str, str, str | None, str | None, str | None]:
     bridge = _bridge_app()
     queue_code = _queue_code_for_queue_id(target_queue_id) or str(fallback_queue_code or "").strip()
     if not queue_code:
@@ -327,7 +327,7 @@ def _resolve_handoff_extension(
             required_skills=required_skills,
         )
         if reserved_agent:
-            return queue_code, reserved_agent["extension"], level, resolved_tenant_id
+            return queue_code, reserved_agent["extension"], level, resolved_tenant_id, reserved_agent["agent_id"]
         raise HTTPException(status_code=409, detail=f"No available {level} agent right now")
 
     extension = bridge._transfer_target_map().get(queue_code)
@@ -336,7 +336,7 @@ def _resolve_handoff_extension(
             status_code=400,
             detail=f"Unknown transfer queue_code: {queue_code}",
         )
-    return queue_code, extension, None, None
+    return queue_code, extension, None, None, None
 
 
 def _resolve_handoff_channel(session, link: AsteriskCallLinkRow) -> str:
@@ -546,11 +546,33 @@ def request_handoff(call_id: str, body: VoiceAIHandoffRequestIn, actor: dict) ->
         if voice_session is None:
             raise HTTPException(status_code=404, detail="Voice AI session not found")
 
-        queue_code, target_extension, resolved_level, resolved_tenant_id = _resolve_handoff_extension(
+        queue_code, target_extension, resolved_level, resolved_tenant_id, reserved_agent_id = _resolve_handoff_extension(
             body.target_queue_id or voice_session.handoff_target_queue_id or link.queue_id,
             fallback_queue_code=_queue_code_for_queue_id(link.queue_id),
             call_id=call_id,
         )
+        escalation: EscalationRow | None = None
+        if reserved_agent_id and target_extension != "7100":
+            now_reserved = utc_now_iso()
+            escalation = EscalationRow(
+                escalation_id=new_id("esc"),
+                call_id=call_id,
+                tenant_id=resolved_tenant_id,
+                from_level=str(link.current_level or "L1"),
+                to_level=resolved_level,
+                reason_code="AI_HANDOFF",
+                required_skills_json="[]",
+                priority=3,
+                status="ringing",
+                real_agent_id=reserved_agent_id,
+                assigned_agent_id=target_extension,
+                attempted_agent_ids_json=json.dumps([reserved_agent_id], ensure_ascii=False),
+                requested_at=now_reserved,
+            )
+            session.add(escalation)
+            session.flush()
+            _append_escalation_timeline(escalation, link, action="escalation.agent_reserved", extra={"agent_id": reserved_agent_id})
+            _emit_escalation_event(event_type="AgentReserved", escalation=escalation, link=link, extra={"agent_id": reserved_agent_id})
         handoff_metadata = body.metadata or {}
         actor_user = str(actor.get("user") or actor.get("sub") or "ai-voice-runtime").strip()
         actor_role = str(actor.get("role") or "admin").strip() or "admin"
@@ -597,15 +619,33 @@ def request_handoff(call_id: str, body: VoiceAIHandoffRequestIn, actor: dict) ->
             target_extension,
             body.target_queue_id or voice_session.handoff_target_queue_id or link.queue_id,
         )
-        ami_result = bridge._ami_action(
-            "Redirect",
-            {
-                "Channel": channel,
-                "Context": bridge._transfer_context(),
-                "Exten": target_extension,
-                "Priority": 1,
-            },
-        )
+        try:
+            ami_result = bridge._ami_action(
+                "Redirect",
+                {
+                    "Channel": channel,
+                    "Context": bridge._transfer_context(),
+                    "Exten": target_extension,
+                    "Priority": 1,
+                },
+            )
+        except Exception as exc:
+            if escalation is not None:
+                try:
+                    routing_release_by_agent_id(reserved_agent_id)
+                except Exception:
+                    LOGGER.warning("bridge.handoff_release_agent_failed call_id=%s agent_id=%s", call_id, reserved_agent_id)
+                escalation.status = "failed"
+                escalation.completed_at = utc_now_iso()
+                session.commit()
+                _append_escalation_timeline(escalation, link, action="escalation.transfer_failed", extra={"agent_id": reserved_agent_id, "error": str(exc)})
+                _emit_escalation_event(event_type="TransferFailed", escalation=escalation, link=link, extra={"agent_id": reserved_agent_id, "error": str(exc)})
+            raise HTTPException(status_code=502, detail=f"Failed to redirect call to agent: {exc}") from exc
+
+        if escalation is not None:
+            set_routing_agent_status(reserved_agent_id, "RINGING")
+            _append_escalation_timeline(escalation, link, action="escalation.agent_ringing", extra={"agent_id": reserved_agent_id})
+            _emit_escalation_event(event_type="AgentRinging", escalation=escalation, link=link, extra={"agent_id": reserved_agent_id})
 
         now = utc_now_iso()
         link.voice_session_id = (