omnius 1.0.606 → 1.0.607

package/README.md

@@ -366,7 +366,8 @@ Context-window dumps are written for main agents, sub-agents, internal runners,
 | Method | Path | Purpose |
 | --- | --- | --- |
 | `GET` | `/v1/voice/state` | Voice runtime status |
-| `GET`/`POST` | `/v1/voice/models` · `/v1/voice/models/switch` | List · switch TTS model |
+| `POST` | `/v1/voice/start` · `/v1/voice/stop` | Enable/warm voice · pause voice input |
+| `GET`/`POST` | `/v1/voice/models` · `/v1/voice/models/switch` | List · switch and enable an exact TTS model |
 | `GET`/`POST` | `/v1/voice/supertonic-settings` | Read · update voice tuning |
 | `GET`/`POST` | `/v1/voice/asr-models` · `/v1/voice/asr-models/switch` | List · switch ASR model |
 | `POST` | `/v1/voice/tts` · `/v1/audio/speech` | Synthesize speech (+ OpenAI alias) |
@@ -379,6 +380,12 @@ Context-window dumps are written for main agents, sub-agents, internal runners,
 | `GET` | `/v1/voicechat/ws` | WebSocket upgrade for full-duplex voicechat |
 | `POST` | `/v1/vision/describe` | Vision describe placeholder |
 
+TTS requests auto-warm the daemon and never silently fall back to a different
+voice. The built-in registry includes GLaDOS, Overwatch, and the packaged
+`luxtts:announcer-testchamber03` clone source. Voicebox defaults to its stable
+sub-model set; use `OMNIUS_VOICEBOX_MODELS=all` for the full suite or a
+comma-separated list of `voicebox_*` IDs for an explicit subset.
+
 ### Generative media
 
 Backed by the unified `~/.omnius` store and shared venvs (see [Shared Media Dependencies](#shared-media-dependencies)). Outputs land in the global gallery at `~/.omnius/media/{images,videos,audio,music}`.