v0.1.0 — Initial scaffold

🎉 First public release

Initial scaffold of Logica Voice — the open-source conversational AI framework.

What's in v0.1

• ✅ 16 packages, TypeScript, pnpm workspaces
• ✅ Core: frame-based pipeline, adapter interfaces, BR phone normalization, mention detection
• ✅ Brain adapters: built-in (YAML agents + any OpenAI-compat LLM) and logicaos (first-class LogicaOS integration)
• ✅ Channel adapters: Telegram (long polling, ACL fail-closed) and WhatsApp (Baileys with LID resolve + BR 9th digit + fail-closed ACL)
• ✅ Voice services: interfaces ready, Python bridge implementation in v0.2
• ✅ CLI: logica-voice init/start/stop/status
• ✅ 9 docs (architecture, stack, decisions, voice, voices, brain-adapters, logicaos-integration, pipecat-comparison, roadmap)

What's coming in v0.2

• Voice services implementation (faster-whisper, Kokoro, F5-TTS via Python bridge)
• Audio in/out in Telegram + WhatsApp
• npx create-logica-voice published to npm
• Docker compose

Quick start

npx create-logica-voice my-bot
cd my-bot
# edit logica-voice.yaml
npm start

MIT license. PRs welcome. No CLA.

🙏 Inspired by Pipecat.