How do I install the Voiceclaw Jp skill?

Run npx skillmds@latest add dvcrn/voiceclaw-jp in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Voiceclaw Jp safe to use?

This skill has not completed SkillMD's automated safety review yet. Independent scanners report: SkillSpector: PASS, Skill Scanner: PASS. Capability flags: executes scripts, makes network calls, reads secrets. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Voiceclaw Jp?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Voiceclaw Jp free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Voiceclaw Jp?

dvcrn (@dvcrn) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Voiceclaw Jp by dvcrn — Agent Skill

Name: Voiceclaw Jp
Rating: 4.5 (4 reviews)
Author: dvcrn

Voiceclaw Jp

Voice conversation interface for OpenClaw using wake word detection, streaming LLM responses, and text-to-speech. Use when a user wants to talk to their OpenClaw agent by voice, set up a voice assistant, or add speech input/output to OpenClaw. Supports configurable wake words, VOICEVOX TTS, and sentence-level streaming for low-latency responses.

dvcrn Updated Sep 8, 2026 32 repo stars

File contents

voiceclaw

Voice conversation skill for OpenClaw: wake word → STT → LLM (streaming) → TTS → playback.

Requirements

OpenClaw running locally (gateway with chatCompletions enabled)
Node.js 18+
VOICEVOX running on localhost:50021 (download)
Chrome/Edge (Web Speech API for STT)
HTTPS for remote mic access (localhost works without HTTPS)

Quick Start

# Install
git clone https://github.com/kentoku24/voiceclaw.git
cd voiceclaw
npm install

# Start (no .env needed if OpenClaw is running locally)
npm start
# → [voiceclaw] OpenClaw config loaded from ~/.openclaw/openclaw.json
# → [voiceclaw] listening on http://127.0.0.1:8788

# Open browser
open http://127.0.0.1:8788

Press 開始, say the wake word (default: アリス), then speak your command.

Configuration

All settings are optional. Set in .env or environment variables:

Variable	Default	Description
`WAKE_WORDS`	アリスちゃん,アリス,...	Comma-separated wake words
`STT_LANG`	ja-JP	Speech recognition language
`OPENCLAW_MODEL`	openclaw	LLM model name
`VOICEVOX_URL`	http://127.0.0.1:50021	VOICEVOX endpoint
`VOICEVOX_SPEAKER`	1	VOICEVOX speaker ID
`HOST`	127.0.0.1	Server bind address
`PORT`	8788	Server port

Gateway token is auto-detected from ~/.openclaw/openclaw.json. Override with OPENCLAW_GATEWAY_TOKEN if needed.

Architecture

Wake word (browser STT) → voiceclaw server → OpenClaw Gateway (streaming)
                                           → sentence-level TTS (VOICEVOX)
                                           → audio playback (Web Audio API)

See docs/architecture.md for the full sequence diagram.

API Endpoints

Method	Path	Description
GET	`/health`	Health check
GET	`/api/config`	Client-safe settings (wake words, STT lang)
POST	`/api/chat-stream`	Streaming LLM → sentence-level SSE
POST	`/api/chat`	Non-streaming LLM (fallback)
POST	`/api/tts`	Text → VOICEVOX → WAV audio

---name: voiceclaw-jpdescription: Voice conversation interface for OpenClaw using wake word detection, streaming LLM responses, and text-to-speech. Use when a user wants to talk to their OpenClaw agent by voice, set up a voice assistant, or add speech input/output to OpenClaw. Supports configurable wake words, VOICEVOX TTS, and sentence-level streaming for low-latency responses.---# voiceclawVoice conversation skill for OpenClaw: wake word → STT → LLM (streaming) → TTS → playback.## Requirements- **OpenClaw** running locally (gateway with `chatCompletions` enabled)- **Node.js** 18+- **VOICEVOX** running on localhost:50021 ([download](https://voicevox.hiroshiba.jp/))- **Chrome/Edge** (Web Speech API for STT)- **HTTPS** for remote mic access (localhost works without HTTPS)## Quick Start```bash# Installgit clone https://github.com/kentoku24/voiceclaw.gitcd voiceclawnpm install# Start (no .env needed if OpenClaw is running locally)npm start# → [voiceclaw] OpenClaw config loaded from ~/.openclaw/openclaw.json# → [voiceclaw] listening on http://127.0.0.1:8788# Open browseropen http://127.0.0.1:8788```Press **開始**, say the wake word (default: **アリス**), then speak your command.## ConfigurationAll settings are optional. Set in `.env` or environment variables:| Variable | Default | Description ||---|---|---|| `WAKE_WORDS` | アリスちゃん,アリス,... | Comma-separated wake words || `STT_LANG` | ja-JP | Speech recognition language || `OPENCLAW_MODEL` | openclaw | LLM model name || `VOICEVOX_URL` | http://127.0.0.1:50021 | VOICEVOX endpoint || `VOICEVOX_SPEAKER` | 1 | VOICEVOX speaker ID || `HOST` | 127.0.0.1 | Server bind address || `PORT` | 8788 | Server port |Gateway token is auto-detected from `~/.openclaw/openclaw.json`. Override with `OPENCLAW_GATEWAY_TOKEN` if needed.## Architecture```Wake word (browser STT) → voiceclaw server → OpenClaw Gateway (streaming) → sentence-level TTS (VOICEVOX) → audio playback (Web Audio API)```See [docs/architecture.md](docs/architecture.md) for the full sequence diagram.## API Endpoints| Method | Path | Description ||---|---|---|| GET | `/health` | Health check || GET | `/api/config` | Client-safe settings (wake words, STT lang) || POST | `/api/chat-stream` | Streaming LLM → sentence-level SSE || POST | `/api/chat` | Non-streaming LLM (fallback) || POST | `/api/tts` | Text → VOICEVOX → WAV audio |

Frequently asked questions