Skip to content

CHANGELOG.md

Latest commit

 

History

History
81 lines (71 loc) · 3.52 KB

CHANGELOG.md

File metadata and controls

81 lines (71 loc) · 3.52 KB
Raw

Changelog

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

0.2.0 - 2026-01-14

Added

  • Dynamic TTS Voice Selection: Choose from 40+ voices for Aura models
    • Aura 2 EN: 40 English voices (luna, mars, zeus, athena, and more)
    • Aura 2 ES: 10 Spanish voices (aquila, sirio, diana, and more)
    • Aura 1: 12 voices (angus, asteria, orion, and more)
  • TTS Language Selection: Select language for MeloTTS (English, Spanish, French, Chinese, Japanese, Korean)
  • Real-time Conversation State Tracking: Monitor exactly what your voice assistant is doing
    • Shows current processing stage (idle, processing, waiting for LLM, generating response)
    • Displays specific tool execution details (e.g., "executing: HassTurnOn on 'kitchen light' (1/2)")
    • Shows device/area/floor being controlled
    • Error state indication
  • Configuration UI improvements with helpful descriptions and disclaimers
  • Validation to prevent incompatible voice/language combinations
  • Improved UI labels and descriptions for better user experience

Fixed

  • Conversation entity availability issue (removed incompatible async_migrate_engine call)
  • Invalid voice/language selections are automatically reset to defaults
  • Incompatible voice/language settings are cleaned up when switching models

Technical Details

  • Added CONF_TTS_VOICE and CONF_TTS_LANGUAGE configuration constants
  • Implemented comprehensive voice and language mappings for all TTS models
  • Enhanced conversation entity with state property and real-time updates
  • Added validation logic in config flow to ensure data consistency
  • Updated translation files with proper labels for all new options

0.1.0 - 2026-01-04

Added

  • Initial release of Cloudflare Workers AI integration for Home Assistant
  • Text-to-Speech (TTS) support with 4 models:
    • Deepgram Aura 2 English
    • Deepgram Aura 2 Spanish
    • Deepgram Aura 1 (Multi-language)
    • MyShell.ai MeloTTS (Multi-language)
  • Speech-to-Text (STT) support with 4 models:
    • OpenAI Whisper
    • OpenAI Whisper Large V3 Turbo
    • OpenAI Whisper Tiny English
    • Deepgram Nova 3
  • Conversation (LLM) support with 4 models (all with function calling):
    • Hermes 2 Pro Mistral 7B
    • Llama 4 Scout 17B
    • Llama 3.3 70B Fast
    • Mistral Small 3.1 24B
  • Full device control via voice with 22 built-in tools:
    • Device control (turn on/off)
    • Light control (color, brightness)
    • Media control (play, pause, next, previous)
    • Volume control
    • Fan control
    • Vacuum control
    • Timer management
    • And more
  • Natural language color interpretation ("set to the color of the sky")
  • Area-based device control ("turn on all lights in the living room")
  • Floor-based device control
  • Domain filtering for targeted control (lights only, switches only, etc.)
  • Configuration flow with model selection
  • Full Home Assistant Assist pipeline integration
  • Comprehensive debug logging for troubleshooting

Technical Details

  • Proper voluptuous schema parsing for Home Assistant intent system
  • Correct parameter mapping for device control tools
  • Enhanced system prompts for accurate entity name usage
  • Comprehensive HTTP/LLM request/response logging
  • Support for Cloudflare Workers AI function calling format