Vibetty

WebSocket terminal server with voice input support and Claude AI intelligent interaction.

Features

• WebSocket Terminal - Real-time terminal web interface based on Axum framework
• Voice Input - Speech-to-text support for executing commands via voice
• Claude AI Integration - AI-assisted terminal interaction using echokit_terminal
• Multiple ASR Support
  • OpenAI Whisper API
  • Alibaba Cloud Paraformer real-time speech recognition (todo)

Quick Start

ASR Configuration

Vibetty supports two speech recognition modes:

Interactive Setup (Recommended)

Run the setup wizard to configure ASR interactively:

vibetty setup

This launches a TUI where you can:

1. Select a platform: Whisper or WebVosk
2. If Whisper, choose a provider preset: OpenAI, ByteFuture, Groq, GLM, or Custom
3. Fill in API key and other settings
4. Configuration is saved to ~/.vibetty/config.toml

Manual Configuration

You can also configure ASR manually via environment variables.

Option 1: Whisper API (Server-side)

Create a .env file and configure the Whisper API (Groq recommended). Alternatively, you can set these environment variables directly in your shell configuration file (e.g., ~/.bashrc or ~/.zshrc):

VIBECODE_ASR_API_KEY=your_api_key_here
VIBECODE_ASR_URL=https://api.groq.com/openai/v1/audio/transcriptions
VIBECODE_ASR_MODEL=whisper-large-v3
VIBECODE_ASR_LANG=en
VIBECODE_ASR_PROMPT=

Then start the service:

Option A: Download pre-built binary

Download the latest release from the releases page:

# After downloading
./vibetty -- claude

Option B: Build from source

# Build the release binary
cargo build --release
./target/release/vibetty -- claude

Tip: To run vibetty from any directory, place the binary in a directory on your PATH. If it exists, we recommend ~/.cargo/bin:

/usr/local/bin:/usr/bin:/bin:/home/user/.cargo/bin

echo $PATH

echo $PATH | grep -q "$HOME/.cargo/bin" && echo "Yes" || echo "No"

# For pre-built binary
mv vibetty ~/.cargo/bin/

# Or for self-compiled binary
mv target/release/vibetty ~/.cargo/bin/

# For pre-built binary
mv vibetty ~/.cargo/bin/

# Or for self-compiled binary
mv target/release/vibetty ~/.cargo/bin/

Option 2: WebVosk (Browser-side)

Speech recognition runs entirely in the browser using Vosk models. No API key required.

Option A: Download pre-built binary

# Set ASR platform and run
VIBECODE_ASR_PLATFORM=web_vosk ./vibetty -- claude

Option B: Build from source

# Set ASR platform and run
VIBECODE_ASR_PLATFORM=web_vosk ./vibetty -- -- claude

Then visit the WebVosk interface at: http://localhost:3000/vosk

Note: First-time use requires downloading Vosk model files (~40MB each). The models are cached in your browser.

For more options, use --help:

./vibetty --help

Visit: http://localhost:3000 after starting the service.

API Reference

Change Directory

Change the current working directory via HTTP API.

Endpoint: POST /api/change-dir

Request:

curl -X POST http://localhost:3000/api/change-dir \
  -H "Content-Type: application/json" \
  -d '{"path": "/path/to/directory"}'

Example:

# Change to absolute path
curl -X POST http://localhost:3000/api/change-dir \
  -H "Content-Type: application/json" \
  -d '{"path": "/home/user/documents"}'

# Change to relative path
curl -X POST http://localhost:3000/api/change-dir \
  -H "Content-Type: application/json" \
  -d '{"path": "../parent-folder"}'

Note: This endpoint only accepts requests from localhost for security reasons.

Environment Variables

Note: Legacy environment variables (e.g. ASR_URL, ASR_API_KEY, VIBETTY_EXIT_COMMAND, etc.) have been renamed with the VIBECODE_ prefix. Using old names will trigger a warning but still work. Please migrate to the new names.

Platform Support

Currently supports Linux and macOS. Windows is not supported because the pty-process library (used for pseudo-terminal handling) is Unix-only and does not support Windows ConPTY.