据 AlphaMoat 统计,截至 2026-09,LiveAvatar 累计下载 3,263 次、获 9 个星标,在全部 5,078 个 AI 技能中排名第 3,488 位,在「设计创意」分类(512 个)中排名第 320 位。
| 平台 | 累计安装 | 累计下载 | 星标 | 评论 |
|---|---|---|---|---|
| ClawHub ↗ | 104 | 3,122 | 9 | 2 |
| SkillHub ↗ | 294 | 3,263 | 9 | — |
各平台的计数口径不同,统计时间不一致。
它启动一个由 LiveAvatar 驱动的实时 AI 头像:你对麦克风说话,头像把语音转成文字发给 OpenClaw Gateway,再把代理的回答以口型同步的视频播报出来。功能包括语音对语音对话、文字聊天兜底、长回答的智能 TTS 摘要、回声消除(不会回应自己)与多种头像选择。使用前需要设置 LIVEAVATAR_API_KEY,并要求 OpenClaw Gateway 在运行、Node.js 18+ 与带麦克风的现代浏览器。
资料发布于 2026-10-27
以下是根据技能文档生成的示例,不代表实际运行结果。
2026-09 起统计方法调整:同名技能按作者分别统计,并覆盖全部技能。图中虚线为调整前的月份,与之后的数据不直接比较,也不计算跨越调整的增长。
| 月份 | 累计下载 | 环比 | 星标 |
|---|---|---|---|
| 2026-09 | 3,263 | — | 9 |
| 2026-08 | 3,041 | +3.3% | 9 |
| 2026-07 | 2,944 | +7.4% | 9 |
| 2026-06 | 2,740 | +5.1% | 9 |
| 2026-05 | 2,607 | — | 9 |
| 2026-03 | 2,251 | — | 9 |
以下是作者发布的原始文件,按纯文本显示,本站未做修改或审核;其中的链接不可点击,命令请核实后再运行。抓取于 2026-10-23。
---
name: liveavatar
description: Talk face-to-face with your OpenClaw agent using a real-time video avatar powered by LiveAvatar
user-invocable: true
metadata: {"openclaw":{"emoji":"🎭","requires":{"env":["LIVEAVATAR_API_KEY"],"bins":["node","npm"]},"install":[{"id":"node","kind":"node","package":"openclaw-liveavatar","bins":["openclaw-liveavatar"],"label":"Install LiveAvatar (npm)"}]}}
---
# OpenClaw LiveAvatar
Give your OpenClaw agent a face and voice! This skill launches a real-time AI avatar that you can talk to naturally using your microphone. The avatar listens to you, sends your speech to your OpenClaw agent, and speaks the response back with lip-synced video.
Powered by [LiveAvatar](https://liveavatar.com) - real-time AI avatar technology.
## Setup
### 1. Get Your API Key (Free)
1. Go to [app.liveavatar.com](https://app.liveavatar.com)
2. Create a free account
3. Copy your API key from the dashboard
### 2. Set Your API Key
```bash
export LIVEAVATAR_API_KEY=your_api_key_here
```
Or add to `~/.openclaw/openclaw.json`:
```json
{
"skills": {
"entries": {
"liveavatar": {
"env": {
"LIVEAVATAR_API_KEY": "your_api_key_here"
}
}
}
}
}
```
## Usage
Run `/liveavatar` to start the video avatar interface.
When the user runs this command:
1. **Check if LIVEAVATAR_API_KEY is set**. If not, tell them:
> You need a LiveAvatar API key. Get one free at https://app.liveavatar.com
> Then set it: `export LIVEAVATAR_API_KEY=your_key`
2. **Launch the interface**:
```bash
npx openclaw-liveavatar
```
3. **Tell the user**:
> Your LiveAvatar interface is starting at http://localhost:3001
> It will connect automatically to your OpenClaw Gateway.
>
> Tips:
> - Allow microphone access when prompted
> - Click the green mic button to speak
> - The avatar will respond with your agent's answers
> - Click the X button to end the session
## How It Works
```
You speak → Avatar transcribes → OpenClaw processes → Avatar speaks response
```
1. **Voice Input**: Speak into your microphone
2. **Transcription**: LiveAvatar converts speech to text
3. **Agent Processing**: Text sent to OpenClaw Gateway (port 18789)
4. **Response**: Agent response returned
5. **Avatar Speech**: Avatar speaks with natural lip-sync
## Features
- Real-time video avatar with expressions
- Voice-to-voice conversations
- Text chat fallback option
- Smart TTS summarization for long responses
- Echo cancellation (won't respond to itself)
- Multiple avatar choices
## Requirements
- OpenClaw Gateway running (`openclaw gateway`)
- LiveAvatar API key
- Modern browser with microphone
- Node.js 18+
## Troubleshooting
**"OpenClaw Disconnected"**
```bash
openclaw gateway
```
**"No avatars available"**
- Check LIVEAVATAR_API_KEY is set correctly
**Voice not working**
- Allow microphone access in browser
- Check system audio settings
## Links
- [LiveAvatar](https://liveavatar.com) - Real-time avatar platform
- [OpenClaw](https://openclaw.ai) - Your personal AI assistant
- [GitHub](https://github.com/eNNNo/openclaw-liveavatar) - Source code
ContextWeave 架构图一键生成
用 Nano Banana Pro 生成或编辑图片的脚本技能。
用 whisper 命令行在本地把音频转写成文字或做翻译的技能。
通过 MCP 服务生成与编辑图表的技能。
面向界面设计的思路与落地建议技能,覆盖布局、交互流程、设计令牌与前端改动。
面向多家视频模型 API 的 AI 视频生成与编排技能。
用 ffmpeg 从视频里取出画面帧,或生成便于快速查看的缩略图。
用本地 Whisper 把音频文件转成文字,模型下载完成后可完全离线运行。
数据月份:2026-09 · 下载/星标/安装数聚合自公开技能注册库(ClawHub、SkillHub),按月更新。口径说明见数据方法。