Auralite

unlisted

by Christian B. B. Houmann

ā˜… 14 starsUpdated 23d ago
View on GitHub

Auralite

GitHub release License

Auralite is an AI-powered voice assistant plugin for Obsidian that transforms how you interact with your notes. Speak naturally to create content, transcribe thoughts, and manage your knowledge base hands-free.

✨ Key Features

  • šŸŽ™ļø Voice-to-Text Transcription: Convert speech to formatted text with multiple OpenAI models
  • šŸ¤– AI Assistant: Create notes, write content, and edit existing notes using voice commands
  • šŸ“ Smart Formatting: Automatic Markdown formatting with Obsidian-style links and LaTeX equations
  • šŸ”‡ Silence Detection: Automatically stop recording after periods of silence
  • šŸŽÆ Context-Aware: AI understands your vault structure and note relationships

šŸš€ Quick Start

Installation

Option 1: BRAT (Recommended)

  1. Install BRAT from Obsidian's Community Plugins
  2. Open command palette → BRAT: Add a beta plugin for testing
  3. Enter chhoumann/auralite
  4. Enable the plugin in Settings → Community Plugins

Option 2: Manual Installation

  1. Download manifest.json, main.js, and styles.css from the latest release
  2. Place files in <vault>/.obsidian/plugins/auralite/
  3. Enable the plugin in Settings → Community Plugins

Setup

  1. Open Obsidian Settings → Auralite
  2. Enter your OpenAI API key
  3. Choose your preferred transcription model (whisper-1, gpt-4o-transcribe, or gpt-4o-mini-transcribe)

šŸŽÆ Core Commands

Voice Transcription

Command: Transcribe

  • Speak your thoughts and get formatted text at your cursor
  • Automatic punctuation and paragraph breaks
  • Supports multiple languages and accents

AI Assistant

Command: Toggle AI Assistant Listening

  • Create Notes: "Create a new note about quantum physics"
  • Write Content: "Write a summary of today's meeting"
  • Edit Text: "Make the second paragraph more concise"
  • Add Sections: "Add a conclusion to this note"

Utility

Command: Cancel Ongoing Operation

  • Stop any AI task in progress

šŸ”§ Transcription Models

Choose the best model for your needs:

ModelBest ForCharacteristics
whisper-1Short English audio, speedFastest, most reliable, open-source
gpt-4o-transcribeNon-English languages, accuracyBetter accuracy, slower, newer model
gpt-4o-mini-transcribeCost-effective transcriptionLightweight, affordable, good for bulk use

šŸ’” Usage Tips

  • Be Specific: Give clear instructions like "Add a bullet list of project goals"
  • Use Context: Reference selected text or current note content
  • Markdown Aware: Ask for formatted content, headers, lists, and equations
  • Math Support: Request LaTeX equations between $$ symbols
  • Template Integration: Set up note templates in settings for consistent formatting

šŸ› ļø Development

Build Commands

# Development with watch mode
bun ./esbuild.config.mjs

# Production build
bun ./esbuild.config.mjs production

# Type checking and linting
bun run check

Project Structure

src/
ā”œā”€ā”€ main.ts              # Plugin entry point
ā”œā”€ā”€ commands.ts          # Obsidian commands
ā”œā”€ā”€ ai.ts               # AI integration
ā”œā”€ā”€ AudioRecorder.ts    # Voice recording
ā”œā”€ā”€ actions/            # AI actions (create, write, edit)
ā”œā”€ā”€ tasks/              # Background tasks
└── components/         # UI components

Contributing

  1. Fork the repository
  2. Create a feature branch: git checkout -b feat/new-feature
  3. Make changes and test thoroughly
  4. Run quality checks: bun run check
  5. Submit a pull request

šŸ“Š Requirements

  • Obsidian: Version 1.6.5 or higher
  • OpenAI API Key: Required for AI features
  • Microphone: For voice input
  • Internet Connection: For AI processing

šŸ“š Documentation

Full documentation is available at auralite.obsidian.guide

šŸ¤ Support

For plugin developers

Search results and similarity scores are powered by semantic analysis of your plugin's README. If your plugin isn't appearing for searches you'd expect, try updating your README to clearly describe your plugin's purpose, features, and use cases.