asd

html/README.md

@@ -6,8 +6,58 @@ Install [Yarn](https://yarnpkg.com/getting-started/install), and run: `yarn inst
 
 ## Development
 
-1. Start ttyd: `ttyd bash`
-2. Start the dev server: `yarn run start`
+1. (Optional) Set OpenAI API key for speech recognition and TTS summarization: `export OPENAI_API_KEY=your_key_here`
+2. Start ttyd: `ttyd bash`
+3. Start the dev server: `yarn run start`
+
+## Speech Recognition
+
+The terminal includes speech-to-text functionality using OpenAI's Whisper API. To enable:
+
+1. Set the `OPENAI_API_KEY` environment variable
+2. Click the microphone button or press `Ctrl+M` to start/stop recording
+3. Speak clearly into your microphone
+4. The transcribed text will be sent to the terminal when recording stops
+
+### Voice Commands
+
+The following voice commands are recognized and converted to key presses:
+
+- **"enter"** or **"return"** → Enter key
+- **"tab"** → Tab key
+- **"escape"** → Escape key
+- **"backspace"** → Backspace key
+- **"delete"** → Delete key
+- **"control c"** or **"ctrl c"** → Ctrl+C
+- **"control d"** or **"ctrl d"** → Ctrl+D
+- **"control z"** or **"ctrl z"** → Ctrl+Z
+- **"control [a-z]"** → Ctrl+[letter] combinations
+
+Any other speech will be typed as regular text.
+
+## Text-to-Speech
+
+The terminal includes intelligent text-to-speech functionality that summarizes and reads terminal output aloud. Features:
+
+1. **Toggle TTS**: Click the speaker button (🔊) or press `Ctrl+T` to enable/disable
+2. **Pause/Resume**: Press `Ctrl+P` to pause/resume speech when active
+3. **Smart Summarization**: Uses OpenAI's API to intelligently summarize terminal output
+4. **Voice Control**: Configurable voice, rate, volume, and pitch settings
+
+### TTS Features
+
+- **AI Summarization**: Uses OpenAI GPT-4o-mini to summarize long terminal output into concise, spoken-friendly summaries
+- **ANSI Filtering**: Removes terminal control sequences for clean speech
+- **Text Buffering**: Intelligently buffers text before speaking to avoid choppy audio
+- **Queue Management**: Handles multiple text chunks with proper timing
+- **Browser Compatibility**: Uses standard Web Speech API (Speech Synthesis)
+- **Fallback Support**: Works even without API key (reads raw text when summarization unavailable)
+
+### Visual Indicators
+
+- **Gray Button**: TTS disabled
+- **Green Button**: TTS enabled
+- **Overlay Messages**: Shows "🔊 TTS Speaking" when active
 
 ## Publish