Package Exports

This package does not declare an exports field, so the exports above have been automatically detected and optimized by JSPM instead. If any package subpath is missing, it is recommended to post an issue to the original package (claude-code-speak) to support the "exports" field. If that is not possible, create a JSPM override to customize the exports field for this package.

Readme

claude-code-speak

Give Claude Code a voice. This plugin adds ElevenLabs text-to-speech to Claude Code, so it can read its responses aloud.

Perfect for people who are new to the CLI and want a friendlier, more human experience with Claude Code.

Install

claude plugin add claude-code-speak

Setup

Run the setup command inside Claude Code:

/speak-setup

You'll need an ElevenLabs API key (free tier available). The setup wizard walks you through picking your key and choosing a voice.

Commands

Command	What it does
`/speak`	Read the last response aloud
`/speak hello!`	Speak specific text
`/speak-on`	Auto-speak every response
`/speak-off`	Turn off auto-speak
`/speak-setup`	Configure API key and voice

How it works

When you trigger /speak, Claude summarizes its last response into natural speech (stripping code, markdown, and symbols) and sends it to ElevenLabs TTS. The audio plays through your system speaker.

Auto-speak mode (/speak-on) does this automatically after every response until you turn it off.

Requirements

Python 3 (pre-installed on macOS/most Linux)
ElevenLabs API key (free at elevenlabs.io)
macOS or Linux (uses afplay on Mac, mpv on Linux)

Configuration

Config is stored in ~/.config/claude-code-speak/:

api_key — your ElevenLabs API key
voice_id — your chosen voice (optional, has a default)
model — TTS model (optional, defaults to eleven_turbo_v2_5)

License

MIT