Add Discord-native hybrid front-end for Jarvis (bot + bridge)

Transform isair/jarvis into a Discord-controlled voice assistant running on
the Ubuntu VNC desktop, keeping the mature ~39k-line Python brain intact.

- bot/ (Node + bun, discord.js): /자비스 slash commands (ephemeral),
  voice channel join + voice receive/playback, pluggable VNC screen broadcast
  (selfbot live / noVNC / screenshot)
- bridge/ (Python, Flask): wraps jarvis STT + run_reply_engine + Piper TTS
  behind a thin localhost HTTP API
- .env.example, scripts/ (start_bridge/start_bot/dev), README rewrite,
  docs/language-comparison.md and docs/vnc-xfce-setup.md

Language decision: hybrid (Python brain + Node/bun Discord layer) because
Discord blocks bot video; native screen broadcast only works via a Node
selfbot library.

requirements.txt

@@ -0,0 +1,40 @@
+python-dotenv==1.0.1
+flask>=3.0.0
+requests==2.32.3
+beautifulsoup4>=4.12.0
+lxml>=4.9.0
+html2text>=2020.1.16
+playwright>=1.40.0
+numpy<2.0.0
+faster-whisper==1.0.3
+setuptools<81
+sounddevice==0.4.7
+pytesseract==0.3.13
+Pillow==10.4.0
+webrtcvad==2.0.10
+rapidfuzz==3.6.1
+pynput>=1.7.6
+geoip2==4.8.0
+tzdata==2026.1; sys_platform == "win32"
+miniupnpc==2.2.8
+pytest==8.3.2
+pytest-repeat==0.9.3
+mcp==1.13.1
+chatterbox-tts==0.1.2
+piper-tts>=1.3.0
+pygame>=2.1.0
+faiss-cpu>=1.7.4
+
+# NVIDIA CUDA libraries for GPU-accelerated speech recognition on Windows
+nvidia-cublas-cu12>=12.8.0; sys_platform == "win32"
+nvidia-cudnn-cu12>=9.0.0; sys_platform == "win32"
+
+# MLX Whisper for Apple Silicon Macs (much faster than CPU-based faster-whisper)
+mlx-whisper>=0.4.0; sys_platform == "darwin" and platform_machine == "arm64"
+
+# Desktop app dependencies
+PyQt6>=6.6.0
+PyQt6-WebEngine>=6.6.0
+psutil>=5.9.0
+# Note: 6.13.0+ has VC runtime pre-loading fix for Windows
+pyinstaller>=6.13.0