create-axiom-body
v2.1.0
Published
Give any AI agent a physical body — eyes, ears, voice, face. Patent Pending. One command install.
Maintainers
Readme
create-axiom-body
Give any AI agent a physical body in one command. Patent Pending.
Install
npx create-axiom-body my-agentThe positional argument is the install directory (defaults to axiom-body).
The installer:
- Checks/installs Python, git, VC++ Redistributable, and ffmpeg
- Clones the AXIOM Body repo
- Installs all Python dependencies (YOLO, Whisper, Silero VAD, etc.)
- Creates config files
- Generates a start script for your OS
What You Get
| Component | What It Does | Tech |
|-----------|-------------|------|
| Eyes | Real-time person/object detection | YOLOv8 + OpenCV |
| Ears | Speech recognition with VAD + hallucination scrubbing | Silero VAD + Whisper |
| Voice | Word-level lip-sync TTS, SAPI fallback | ElevenLabs + Windows SAPI |
| Face | 10-mode animated face, sense-toggle overlays | HTML5 Canvas / BLE |
| Persona | 33-agent catalog + 2.83 × 10¹⁹ component combinations | personas/catalog/ |
| Brain Bridge | Connect any LLM to the body | File-based IPC |
Requirements
- Python 3.10+
- Node.js 18+
- Git
- ffmpeg (optional, for voice)
- ElevenLabs API key (optional, for voice)
Quick Start
npx create-axiom-body my-agent
cd my-agent
# Required — seed your agent's face & persona:
python scripts/first-run-onboard.py --name "MyAgent"
# Windows:
start.bat
# Mac/Linux:
./start.shLinks
- GitHub
- QIS Protocol
- YonderClaw — Full AI agent platform
MIT License — Yonder Zenith LLC
