v1.3
feat: replace English fallback sounds with native TTS-generated voice prompts
docs: add TTS attribution and sound text mapping