feat: air-gapped deployment guide, no implicit downloads

- Ship tiktoken's cl100k_base inside the package and build the encoding
  from it, so token counting never downloads anything.
- Default EMBEDDINGS_CACHE_DIR to <data home>/models instead of FastEmbed's
  temp dir, and read tokenizer.json and repo metadata from that cache, so
  a model downloads once and survives reboots.
- TTS_PROVIDER=none and STT_PROVIDER=none switch the speech features off:
  the endpoints return 404, audio files fail to ingest with a clear
  message, /api/config reports tts_available/stt_available, and the UI
  hides the Speak and microphone buttons.
- Drop the Google Fonts Roboto import from the web UI.
- prefetch-models fills the cache the app reads; verify-offline checks the
  packaged encoding.
- Docs: new Air-Gapped Deployment guide, settings and cache notes.

frontend/src/components/MessageInput.tsx

@@ -30,6 +30,7 @@ import {
   selectSelectedDocs,
   selectSelectedModel,
   selectSourceDocs,
+  selectSttAvailable,
   selectToken,
   setSelectedDocs,
 } from '../preferences/preferenceSlice';
@@ -348,6 +349,7 @@ export default function MessageInput({
   const token = useSelector(selectToken);
   const attachments = useSelector(selectAttachments);
   const selectedModel = useSelector(selectSelectedModel);
+  const sttAvailable = useSelector(selectSttAvailable);
   const unreadableAttachmentIds = useMemo(
     () =>
       new Set(
@@ -1771,7 +1773,7 @@ export default function MessageInput({
                 loading={toolsLoading}
               />
             )}
-            {ENABLE_VOICE_INPUT && (
+            {ENABLE_VOICE_INPUT && sttAvailable && (
               <MicButton
                 recordingState={recordingState}
                 loading={loading}