feat: add CAMB AI audio toolkit and core TTS provider

[neilruaro-camb]

Summary

• Adds a new CAMB AI audio toolkit tool (camb_ai_audio) to the music_audio toolkit, exposing TTS, transcription, translation, translated TTS, voice cloning, text-to-sound, audio separation, and dubbing capabilities via the CAMB AI API
• Adds a CAMB AI core TTS synthesizer (camb-ai-synthesizer.ts) so Leon can use CAMB AI as its native text-to-speech provider (LEON_TTS_PROVIDER=camb-ai)
• Updates SHOULD_START_PYTHON_TCP_SERVER to only require the Python TCP server when using local STT/TTS providers, allowing external providers like CAMB AI to work without it
• Includes an integration test (test_camb_ai_leon.py) that validates all 8 CAMB AI API endpoints