Skip to content

v0.0.19

Choose a tag to compare

View all tags
@RikyZ90 RikyZ90 released this 09 Apr 18:30
v0.019
09a8a81

[0.0.19] - 2026-04-09

Added

  • Agent Settings UI
    • Model input field now has history tracking and auto-completion from previously used models.
    • Provider input field changed to a dropdown showing only configured providers (API key, local base URL, or OAuth), defaulting to "auto".
  • Audio Messaging Support (STT & TTS)
    • Integrated multi-provider Speech-to-Text (STT) pipeline using OpenAI-compatible APIs (e.g., Groq/Whisper).
    • Browser-native Text-to-Speech (TTS) for agent responses with automatic markdown/code cleaning.
    • Automatic Voice Activity Detection (VAD) with silence threshold and duration settings.
  • WebUI Enhancements
    • High-quality visual feedback for voice recording with pulse animation on the microphone button.
    • Transcription feedback: "Transcribing..." placeholder with shimmer effect during audio processing.
    • Dedicated "Voice & Audio" section in Agent Settings to configure provider URL, API key, and model.
    • TTS user preference persistence via localStorage.
  • Backend Improvements
    • New AudioConfig schema for central management of speech settings.
    • Refactored transcribe_audio Socket.IO event handler for better performance and reliability.

Changed

  • UI Refinements
    • Improved chat input bar aesthetics: microphone and attachment (clip) buttons are now closer and visually aligned.
    • Text-to-Speech (Bot Voice) now defaults to "off" for a cleaner initial experience.

Fixed

  • Code Hygiene
    • Removed unused properties and redundant comments in speech and socket modules.
    • Refactored backend imports and improved error handling for transcription failures.

Full Changelog: v0.0.18...v0.019

Assets 3
Loading