voice-vibecoder 2.18.0__tar.gz → 2.19.0__tar.gz

{voice_vibecoder-2.18.0 → voice_vibecoder-2.19.0}/CHANGELOG.md

@@ -2,6 +2,14 @@
 
 <!-- version list -->
 
+## v2.19.0 (2026-02-23)
+
+### Features
+
+- Replace audio opt-in with text_only flag in sync_state protocol
+  ([`5c428ab`](https://github.com/snokam/voice-vibecoder/commit/5c428ab2c231e91b72fd72dfc3eece6597486ffe))
+
+
 ## v2.18.0 (2026-02-23)
 
 ### Features

{voice_vibecoder-2.18.0 → voice_vibecoder-2.19.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: voice-vibecoder
-Version: 2.18.0
+Version: 2.19.0
 Summary: Olaf The Vibecoder — a voice-controlled coding assistant using OpenAI Realtime API + Claude Code
 Project-URL: Homepage, https://github.com/snokam/voice-vibecoder
 Project-URL: Repository, https://github.com/snokam/voice-vibecoder

{voice_vibecoder-2.18.0 → voice_vibecoder-2.19.0}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "hatchling.build"
 
 [project]
 name = "voice-vibecoder"
-version = "2.18.0"
+version = "2.19.0"
 description = "Olaf The Vibecoder — a voice-controlled coding assistant using OpenAI Realtime API + Claude Code"
 readme = "README.md"
 requires-python = ">=3.10"

{voice_vibecoder-2.18.0 → voice_vibecoder-2.19.0}/src/voice_vibecoder/__init__.py

@@ -3,7 +3,7 @@
 A voice-controlled coding assistant using OpenAI Realtime API + Claude Code.
 """
 
-__version__ = "2.18.0"
+__version__ = "2.19.0"
 
 from voice_vibecoder.app_config import VoiceConfig
 from voice_vibecoder.ui.screen import VoiceCodingScreen

{voice_vibecoder-2.18.0 → voice_vibecoder-2.19.0}/src/voice_vibecoder/server.py

@@ -6,9 +6,11 @@ The browser sends/receives JSON frames — never talks to OpenAI directly.
 
 Protocol (client -> server):
   {"type": "auth", "token": "<JWT>"}           — must be the first message when auth is enabled
-  {"type": "sync_state", "audio": true, "ui_commands": true}  — full state sync (audio/ui_commands opt in)
+  {"type": "sync_state", "ui_commands": true}   — full state sync, connects voice provider
+  {"type": "sync_state", "text_only": true}     — text-only mode: skip voice provider entirely
   {"type": "audio", "data": "<base64 PCM16 24kHz>"}
   {"type": "audio.commit"}
+  {"type": "text", "text": "..."}               — send text (routes via voice or directly to agent)
   {"type": "update_settings", "settings": {...}}  — partial settings update, reconnects provider
 
 Protocol (server -> client):
@@ -110,7 +112,6 @@ class WebSocketBridge:
         self._user_jwt = user_jwt
         self._messages_task: asyncio.Task | None = None
         self._ui_commands_enabled = False
-        self._audio_enabled = False
 
     async def run(self) -> None:
         provider = create_provider(self._settings)
@@ -138,8 +139,8 @@ class WebSocketBridge:
         # Load saved instances — panels appear dynamically as sessions/instances are created
         restored = self._session.registry.load_state()
 
-        # Don't connect voice provider yet — wait for sync_state with audio: true.
-        # Text-only clients skip the voice provider entirely.
+        # Don't connect voice provider yet — wait for sync_state.
+        # Text-only clients (no sync_state) skip the voice provider entirely.
         self._send({"type": "status", "status": "ready"})
 
         try:
@@ -190,12 +191,10 @@ class WebSocketBridge:
                 elif msg_type == "update_settings":
                     await self._handle_settings_update(msg.get("settings", {}))
                 elif msg_type == "sync_state":
-                    # Opt-in flags
                     if msg.get("ui_commands"):
                         self._ui_commands_enabled = True
-                    if msg.get("audio"):
-                        self._audio_enabled = True
-                        # Connect voice provider now (lazy)
+                    if not msg.get("text_only"):
+                        # Connect voice provider (skipped for text_only clients)
                         await self._connect_voice_provider()
                     self._send({"type": "status", "status": "syncing"})
                     await self._sync_state_to_frontend()
@@ -315,7 +314,7 @@ class WebSocketBridge:
             logger.warning("Failed to send WS frame", exc_info=True)
 
     def _on_audio(self, base64_chunk: str) -> None:
-        if not self._audio_enabled:
+        if self._session and self._session.text_only:
             return
         self._send({"type": "audio", "data": base64_chunk})