feat: add voice studio prototype flow

backend/app/schemas/voice_session_schemas.py

@@ -25,6 +25,15 @@ class VoiceTurnCreateFallbackRequest(BaseModel):
     duration_ms: int | None = Field(default=None, ge=1, le=MAX_VOICE_TURN_DURATION_MS)
 
 
+class VoiceTurnUploadAcceptedResponse(BaseModel):
+    """Accepted response for one uploaded-audio voice turn."""
+
+    turn_id: str
+    session_id: str
+    status: str
+    transcription_provider: str | None = None
+
+
 class VoiceSessionFinalizeRequest(BaseModel):
     """Finalize one voice session into a persisted story."""
 
@@ -61,11 +70,14 @@ class VoiceTurnSummaryResponse(BaseModel):
     status: str
     user_transcript: str | None = None
     transcript_confidence: float | None = None
+    transcription_provider: str | None = None
     detected_intent: str
     intent_confidence: float | None = None
     assistant_text: str | None = None
     assistant_audio_ready: bool = False
     assistant_audio_url: str | None = None
+    user_audio_ready: bool = False
+    user_audio_url: str | None = None
     error_message: str | None = None
     created_at: datetime
     updated_at: datetime
@@ -81,10 +93,14 @@ class VoiceSessionSummaryResponse(BaseModel):
     target_mode: str
     status: str
     current_turn_index: int
+    total_turns: int = 0
     working_title: str | None = None
     story_state: dict[str, Any] = Field(default_factory=dict)
     latest_user_transcript: str | None = None
     latest_assistant_text: str | None = None
+    latest_detected_intent: str | None = None
+    latest_assistant_audio_ready: bool = False
+    last_turn_status: str | None = None
     can_continue: bool = False
     can_finalize: bool = False
     last_error: str | None = None