Clawatar provides your AI agent with a fully interactive 3D VRM avatar body featuring lip-sync, expressions, and over 160 animations.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install clawatar
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install clawatar using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Clawatar?
Clawatar is a powerful visual interface designed to bring AI agents to life through high-fidelity 3D models. By utilizing the industry-standard VRM format, this entry in the Openclaw Skills ecosystem allows developers to bridge the gap between text-based interactions and immersive visual experiences. It includes a web-based viewer that handles real-time rendering, skeletal animations, and automatic lip-syncing, transforming a standard LLM into a visible, speaking companion.
The system operates via a local WebSocket server, making it highly extensible and easy to control from any external script or agent. Whether you are building a virtual assistant, a VTuber-style character, or an interactive educational guide, Clawatar provides the technical foundation for realistic 3D representation and voice-driven interaction.
Clawatar Use Cases
- Creating a visual persona for AI-driven customer service bots or personal assistants.
- Building interactive VTuber-style characters controllable via an AI agent.
- Implementing a standalone 3D VRM viewer with integrated voice chat capabilities.
- Developing immersive AI companions for gaming or educational environments.
- Prototyping multi-modal interfaces that combine text, voice, and 3D movement using Openclaw Skills.
How Clawatar Works
- The user launches the Clawatar web viewer which establishes a local WebSocket server at ws://localhost:8765.
- An AI agent or script sends structured JSON commands (such as play_action or speak) to the viewer via the WebSocket protocol.
- The viewer processes the command, triggering specific animations from a pre-loaded library or updating the character facial expressions.
- When a speak command is issued, the system utilizes ElevenLabs for high-quality TTS and synchronizes the avatar lip movements with the audio output.
- The interface provides real-time feedback, allowing users to interact directly via touch reactions or emotion bar controls on the web page.
Clawatar Setup
To get started with this project among your other Openclaw Skills, run the following commands:
# Clone and install
git clone https://github.com/Dongping-Chen/Clawatar.git ~/.openclaw/workspace/clawatar
cd ~/.openclaw/workspace/clawatar && npm install
# Start the Vite development server and WebSocket bridge
npm run start
Once running, navigate to http://localhost:3000. You will need to provide your own .vrm model by dragging and dropping it onto the page or configuring the model.url in clawatar.config.json. For voice features, ensure your ElevenLabs API key is configured in your environment variables.
Clawatar Data Schema & Taxonomy
Clawatar organizes its assets and configuration to ensure compatibility with various Openclaw Skills workflows:
| Component | Details |
|---|---|
| VRM Model | User-provided 3D model file (.vrm). |
| Animations | A library of 162 unique motion files stored in public/animations/catalog.json. |
| Configuration | clawatar.config.json manages ports, default model paths, and voice settings. |
| WebSocket API | Accepts JSON payloads with keys like type, action_id, name, weight, and text. |
| Expressions | Built-in support for happy, angry, sad, surprised, and relaxed. |
Clawatar Advanced Features
- Real-time lip-syncing driven by advanced TTS analysis for natural speech visualization.
- Extensive animation library including 162 distinct motions for greetings, thinking, dancing, and more.
- Multi-scene support with environment presets like Sakura Garden, Night Sky, and Sunset.
- Dynamic camera presets for Face, Portrait, and Full Body shots to enhance cinematic presentation.
- Low-latency WebSocket control, allowing for seamless integration into complex Openclaw Skills automation pipelines.
SKILL.md
Loading
Related Openclaw Skills
Notion API 2026 Manager
A comprehensive toolset for programmatically managing Notion workspaces using the latest 2026 API features.
Xiaochang AI Coding (Teaching Version)
A specialized skill for managing the end-to-end development lifecycle on the Xiaochang AI Coding platform, from requirement analysis to final deployment.
ZeeLin Liberal Arts Paper
A professional AI-driven writing assistant designed to automate the full lifecycle of liberal arts and social science academic paper creation.
ZeeLin Academic Paper Assistant
An AI-driven writing assistant designed for STEM researchers to automate the creation of logically consistent academic papers from title to conclusion.
ClawMe Browser Interaction Skill
A powerful browser automation skill that lets AI agents perform actions in your live Chrome session with manual user confirmation.
Mac AI Optimizer
A comprehensive toolset to transform macOS into a high-performance AI server node by reclaiming system resources and optimizing background processes.
Featured*
