← Experts

Build voice AI apps end to end

Go from zero to a production voice AI app: wire up speech-to-text, route through Claude for reasoning, and synthesize responses with natural-sounding voice output. Build phone agents, voice assistants, real-time translators, and accessibility tools — all without a PhD in audio engineering.

1

Expert Install

Copy this to your agent — it will install, configure, and verify everything.

Say to your agent
Read https://clawhub.md/expert/voice-expert.md and set me up as Voice Agent Builder

Works on OpenClaw, Claude Code, Telegram, Feishu, and any agent interface. Your agent reads voice-expert.md and follows the setup steps inside.

2

Talk and handle it

After setup, say these to your agent:

The reasoning brain of your voice… Official

  • › "Build a voice assistant that books calendar events"
  • › "Create a phone agent that answers customer questions"
  • › "Add voice input/output to my existing Claude app"

Scaffold the voice pipeline — WebSocket… Official

  • › "Build a WebSocket server for real-time audio streaming"
  • › "Add voice activity detection to my app"
  • › "Integrate ElevenLabs TTS into my voice pipeline"

Expose voice capabilities as MCP tools… Official

  • › "Build an MCP tool that triggers a phone call"
  • › "Expose my TTS endpoint as an MCP tool for other agents"
  • › "Create a transcription MCP server for my audio files"
▶ View setup file voice-expert.md

Voice Agent Builder

Agentic setup file — share this URL with your agent and it will set everything up for you: https://clawhub.md/expert/voice-expert.md

Goal: Build voice AI apps end to end

What you'll have: Go from zero to a production voice AI app: wire up speech-to-text, route through Claude for reasoning, and synthesize responses with natural-sounding voice output. Build phone agents, voice assistants, real-time translators, and accessibility tools — all without a PhD in audio engineering.

Step 1: Install

clawhub install anthropics/claude-api openclaw/coding-agent anthropics/mcp-builder

Step 2: Configure

Each skill may need credentials or auth before it can act on your behalf.

openclaw/coding-agent

Scaffold the voice pipeline — WebSocket handlers for streaming audio, VAD (voice activity detection), chunked transcription, and TTS integration.

  • Ensure Claude Code is installed: npm install -g @anthropic-ai/claude-code (or see https://claude.ai/code)
  • Optional — for Codex delegation: npm install -g @openai/codex then codex login
  • No extra environment variables needed if Claude Code is already working in your session

Step 3: Try it

After setup, say these to your agent to verify everything works:

anthropics/claude-api

  • "Build a voice assistant that books calendar events"
  • "Create a phone agent that answers customer questions"
  • "Add voice input/output to my existing Claude app"

openclaw/coding-agent

  • "Build a WebSocket server for real-time audio streaming"
  • "Add voice activity detection to my app"
  • "Integrate ElevenLabs TTS into my voice pipeline"

anthropics/mcp-builder

  • "Build an MCP tool that triggers a phone call"
  • "Expose my TTS endpoint as an MCP tool for other agents"
  • "Create a transcription MCP server for my audio files"

Voice Agent Builder · clawhub.md/expert/voice-expert

← Experts Browse all skills →