feat(voice): first-class voice input and local TTS across web, desktop, and mobile (#2018)

Complete rebuild of voice input on a server-authoritative streaming
architecture, replacing the legacy Web Speech / whole-blob / WASM engines
and the dead voice-agent layer (~4k lines removed).

Speech-to-text (dictation):
- Client streams 16 kHz mono PCM16 chunks over /api/dictation/ws with
  seq/ack ordering; buffered audio is retained and replayed on reconnect
- Server transcribes and streams live partial transcripts back;
  segments auto-commit every ~15s with silence suppression and adaptive
  finalization timeouts
- Local provider (default, zero config): sherpa-onnx models in a forked
  worker process — auto-download with progress, staged extraction with
  verification, corrupt-model auto-recovery, idle shutdown after 5 min
- Model catalog with settings picker (accuracy/speed ratings, sizes,
  download/delete): Parakeet TDT v2 (English) and v3 (25 European
  languages, auto-detected), Whisper base and tiny (multilingual, light)
- OpenAI-compatible provider for any Whisper endpoint
- Composer overlay with live transcript, volume meter, timer, and
  cancel / insert / insert-and-send actions; failed transcriptions keep
  their audio for retry or accepting the partial text as-is
- Configurable keyboard shortcut (default mod+alt+v) toggles dictation;
  Enter confirms and Escape cancels while recording
- Overlay is pixel-aligned with the composer (measured footer height,
  matching paddings/typography/gaps) — no layout shift when toggling

Text-to-speech:
- Local Kokoro provider (English, 11 voices) synthesized in the same
  worker via /api/dictation/tts/speak, managed by the shared model
  pipeline; sentence-pipelined playback keeps time-to-first-audio at
  ~1 sentence regardless of message length, and stop cancels in-flight
  synthesis
- Sanitizer keeps inline-code content (strips backticks only), reads
  interword slashes aloud, and removes only absolute file paths

Settings:
- Voice page unified: a single read-aloud toggle owns all playback
  options (the confusing "Enable Voice Mode" is gone); a new "Enable
  voice input" toggle (default on, persisted to settings.json) hides
  the composer mic entirely when disabled

Mobile and transport:
- iOS/Android microphone permissions added (dictation was previously
  impossible on mobile)
- Fixed Android WebSocket upgrades: the Capacitor WebView origin
  (https://localhost) was missing from the packaged-client allowlist,
  403-ing every WS connection — root cause of the old mobile SSE lock,
  which is now removed for all transports

Security and conventions:
- All HTTP routes sit behind the global /api auth gate; the WS upgrade
  explicitly validates the UI session and origin, with oc_url_token
  narrowly allowlisted and covered by tests; the dictation socket mints
  a fresh URL token before connecting
- Routes register before the generic OpenCode proxy; the client goes
  through runtimeFetch/getRuntimeUrlResolver, and runtime switches
  reset the dictation socket
- VS Code deliberately reports dictation as unavailable (no server
  process in that runtime)

CI: workflow Node bumped 20 -> 22 to match the repo engines and fix
better-sqlite3 installs broken by node-gyp@latest on Node 20.

New dependency: sherpa-onnx-node (prebuilt N-API; macOS/Linux x64+arm64,
Windows x64 — Windows-on-ARM falls back to the OpenAI-compatible provider)

packages/ui/src/App.tsx

@@ -44,7 +44,6 @@ import { ConfigUpdateOverlay } from '@/components/ui/ConfigUpdateOverlay';
 import { AboutDialog } from '@/components/ui/AboutDialog';
 import { RuntimeAPIProvider } from '@/contexts/RuntimeAPIProvider';
 import { registerRuntimeAPIs } from '@/contexts/runtimeAPIRegistry';
-import { VoiceProvider } from '@/components/voice';
 import { useUIStore } from '@/stores/useUIStore';
 import { useGitHubAuthStore } from '@/stores/useGitHubAuthStore';
 import { useFeatureFlagsStore } from '@/stores/useFeatureFlagsStore';
@@ -928,8 +927,8 @@ function App({ apis }: AppProps) {
   }
 
   // Always mount the full provider tree to avoid remounts when isInitialized
-  // flips from false → true. FireworksProvider and VoiceProvider are lightweight
-  // shells; their heavy children are only activated when actually needed.
+  // flips from false → true. FireworksProvider is a lightweight shell; its
+  // heavy children are only activated when actually needed.
   const isBootShell = !isInitialized && !isDesktopRuntime;
 
   return (
@@ -937,7 +936,6 @@ function App({ apis }: AppProps) {
       <SyncProvider key={runtimeEndpointEpoch} sdk={opencodeClient.getSdkClient()} directory={currentDirectory || ''}>
         <RuntimeAPIProvider apis={apis}>
           <FireworksProvider>
-            <VoiceProvider>
               <TooltipProvider delayDuration={300} skipDelayDuration={150}>
                 <div className={isDesktopRuntime ? 'h-full text-foreground bg-transparent' : 'h-full text-foreground bg-background'}>
                   <SyncAppEffects embeddedBackgroundWorkEnabled={embeddedBackgroundWorkEnabled} />
@@ -955,7 +953,6 @@ function App({ apis }: AppProps) {
                   )}
                 </div>
               </TooltipProvider>
-            </VoiceProvider>
           </FireworksProvider>
         </RuntimeAPIProvider>
       </SyncProvider>