/openclaw-voice
Browser-based voice chat for OpenClaw agents.
$ npx -y skills add Purple-Horizons/openclaw-voice --skill openclaw-voice --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
- Fires itselfAuto-invocation. Claude auto-loads it when your prompt matches the work.Auto-invocation is when the right skill fires by itself at the right moment, driven by a FLOW.md router and a hook, instead of you invoking it by name. It is the difference between a skill being installed and a skill actually getting used.Read the full definition →
- You can call itInvoke it directly when you want it.
- Slash command
/openclaw-voice
Context preview
The summary Claude sees to decide when to auto-load this skill.
Browser-based voice chat for OpenClaw agents.
SKILL.md
openclaw-voice.SKILL.mdOpenClaw Voice Skill
Browser-based voice chat for OpenClaw agents.
What It Does
Adds voice chat capability to your OpenClaw agent. Users can speak to your agent via a web browser and hear responses in real-time.
Stack
- **STT**: faster-whisper (local, no API costs)
- **TTS**: ElevenLabs (cloud, high quality) or Chatterbox (self-hosted)
- **Transport**: WebSocket
- **Backend**: OpenClaw gateway (chatCompletions endpoint)
Quick Setup
1. Enable Gateway Endpoint
Add to your `openclaw.json`:
{
"gateway": {
"http": {
"endpoints": {
"chatCompletions": {
"enabled": true
}
}
}
}
}2. Add Voice Agent
{
"agents": {
"list": [
{
"id": "voice",
"workspace": "/path/to/your/workspace",
"model": "anthropic/claude-sonnet-4-5"
}
]
}
}3. Run the Server
# Clone and setup
git clone https://github.com/Purple-Horizons/openclaw-voice.git
cd openclaw-voice
uv sync # or pip install -r requirements.txt
# Configure
cp .env.example .env
# Edit .env with your OPENCLAW_GATEWAY_URL, OPENCLAW_GATEWAY_TOKEN, ELEVENLABS_API_KEY
# Run
PYTHONPATH=. python -m src.server.main
4. Access
Open `http://localhost:8765` in your browser.
For HTTPS (required for mobile mic), use Tailscale Funnel or your own SSL.
Environment Variables
| Variable | Required | Description | |----------|----------|-------------| | `OPENCLAW_GATEWAY_URL` | Yes* | OpenClaw gateway URL (e.g., `http://localhost:18789`) | | `OPENCLAW_GATEWAY_TOKEN` | Yes* | Gateway auth token | | `ELEVENLABS_API_KEY` | Recommended | For high-quality TTS | | `OPENAI_API_KEY` | Fallback | Used if gateway not configured |
*Required for full agent integration. Falls back to direct OpenAI if not set.
Features
- **Push-to-talk**: Hold button to speak
- **Continuous mode**: Hands-free, auto-listens after responses
- **Keyboard shortcut**: Spacebar toggles recording
- **Mobile support**: Works on phones (requires HTTPS)
API Key Auth (Optional)
For production, enable auth:
OPENCLAW_REQUIRE_AUTH=true
OPENCLAW_MASTER_KEY=your-secret-key
Then generate user keys via `POST /api/keys`.
Files
- `src/server/main.py` - FastAPI server
- `src/server/stt.py` - Speech-to-text (Whisper)
- `src/server/tts.py` - Text-to-speech (ElevenLabs/Chatterbox)
- `src/server/backend.py` - AI backend (gateway or OpenAI)
- `src/client/index.html` - Browser UI
License
MIT — Built by [Purple Horizons](https://purplehorizons.io)
Read more
OpenClaw Voice Skill
Browser-based voice chat for OpenClaw agents.
What It Does
Adds voice chat capability to your OpenClaw agent. Users can speak to your agent via a web browser and hear responses in real-time.
Stack
- **STT**: faster-whisper (local, no API costs)
- **TTS**: ElevenLabs (cloud, high quality) or Chatterbox (self-hosted)
- **Transport**: WebSocket
- **Backend**: OpenClaw gateway (chatCompletions endpoint)
Quick Setup
1. Enable Gateway Endpoint
Add to your `openclaw.json`:
{
"gateway": {
"http": {
"endpoints": {
"chatCompletions": {
"enabled": true
}
}
}
}
}2. Add Voice Agent
{
"agents": {
"list": [
{
"id": "voice",
"workspace": "/path/to/your/workspace",
"model": "anthropic/claude-sonnet-4-5"
}
]
}
}3. Run the Server
# Clone and setup git clone https://github.com/Purple-Horizons/openclaw-voice.git cd openclaw-voice uv sync # or pip install -r requirements.txt # Configure cp .env.example .env # Edit .env with your OPENCLAW_GATEWAY_URL, OPENCLAW_GATEWAY_TOKEN, ELEVENLABS_API_KEY # Run PYTHONPATH=. python -m src.server.main
4. Access
Open `http://localhost:8765` in your browser.
For HTTPS (required for mobile mic), use Tailscale Funnel or your own SSL.
Environment Variables
| Variable | Required | Description | |----------|----------|-------------| | `OPENCLAW_GATEWAY_URL` | Yes* | OpenClaw gateway URL (e.g., `http://localhost:18789`) | | `OPENCLAW_GATEWAY_TOKEN` | Yes* | Gateway auth token | | `ELEVENLABS_API_KEY` | Recommended | For high-quality TTS | | `OPENAI_API_KEY` | Fallback | Used if gateway not configured |
*Required for full agent integration. Falls back to direct OpenAI if not set.
Features
- **Push-to-talk**: Hold button to speak
- **Continuous mode**: Hands-free, auto-listens after responses
- **Keyboard shortcut**: Spacebar toggles recording
- **Mobile support**: Works on phones (requires HTTPS)
API Key Auth (Optional)
For production, enable auth:
OPENCLAW_REQUIRE_AUTH=true OPENCLAW_MASTER_KEY=your-secret-key
Then generate user keys via `POST /api/keys`.
Files
- `src/server/main.py` - FastAPI server
- `src/server/stt.py` - Speech-to-text (Whisper)
- `src/server/tts.py` - Text-to-speech (ElevenLabs/Chatterbox)
- `src/server/backend.py` - AI backend (gateway or OpenAI)
- `src/client/index.html` - Browser UI
License
MIT — Built by [Purple Horizons](https://purplehorizons.io)
Open-source browser-based voice interface for AI assistants. Talk to your AI like you talk to Alexa — but self-hosted, private, and connected to your own agent. 🌐 Website: openclawvoice.com

