tts.api 0.3.0

• Add source argument to tts(): run chatterbox in-process via the R package (source = "package") or over HTTP (source = "api", the default), with "auto" picking the package when installed
• Call-record sidecars for tts(), speech_clone(), and speech_design()
• Fix voice listing for the /v1/audio/voices endpoint

tts.api 0.2.0

• Rename speech() to tts() for API consistency
• Add qwen3-tts backend: instructions mapping and language parameter
• Add voice library for ~/.cornball/voices/: voice_library(), voice_file(), voice_ensure()

tts.api 0.1.0

• Initial release
• Support for OpenAI-compatible text-to-speech APIs
• Multiple backend support: OpenAI, Chatterbox, ElevenLabs
• Voice cloning support via voice_upload() and speech_clone()
• Speed adjustment via ffmpeg post-processing for local backends