🎯 Features

  • 📚 Complete SDK Documentation - Movement control, sensors, AI integration (240 lines, optimized)
  • 🔌 REST API Reference - 25+ FastAPI endpoints with detailed examples (500+ lines)
  • 📋 OpenAPI Schema - v3.1.0 specification for auto-generating clients in 50+ languages
  • 🌐 Multi-Language Support - Python, JavaScript, Go, Rust, and more via OpenAPI
  • 🤖 AI Integration - LLMs, vision models, multimodal applications, HuggingFace deployment
  • 💡 Production-Ready Examples - Real code samples and patterns
  • ⚡️ Optimized for Claude - Following official Anthropic guidelines (-50% token usage)

📦 Quick Start

Option 1: Download Pre-built Skill (Recommended)

  1. Go to Releases
  2. Download reachy-mini-sdk.skill
  3. In Claude.ai: Settings → Skills → Upload Skill
  4. Select the downloaded file

Option 2: Build from Source

# Clone repository
git clone https://github.com/jjmartres/reachy-mini-sdk-skill.git
cd reachy-mini-sdk-skill

# Initialize environment
make install

# Package the skill
make package

# Upload dist/reachy-mini-sdk.skill to Claude.ai

🚀 Usage with Claude

Example Prompts

Python SDK:

Write code to make Reachy Mini wave hello and nod twice

REST API:

Show me how to control Reachy Mini from JavaScript using the REST API

Web Dashboard:

Create a React dashboard to monitor robot status in real-time

AI Integration:

Build an object detection app that makes Reachy wave when it sees a person

OpenAPI Client Generation

Generate type-safe clients automatically:

# Python client
openapi-generator-cli generate \
  -i skill/references/openapi_schema.json \
  -g python \
  -o reachy-python-client/

# TypeScript types
openapi-typescript skill/references/openapi_schema.json -o reachy-types.ts

# Go client
openapi-generator-cli generate \
  -i skill/references/openapi_schema.json \
  -g go \
  -o reachy-go-client/

50+ languages supported!

📚 Documentation

Skill Contents

File Description Lines
SKILL.md Core guide (optimized) 240
installation.md Setup for all platforms 150
movement_control.md Complete movement guide 450+
sensors.md Camera, microphone, IMU 200
ai_integration.md AI models and apps 300
daemon_api.md REST API reference 500+
openapi_schema.json OpenAPI v3.1.0 spec Full
openapi_usage.md Client generation guide 200
api_quick_reference.md Quick lookup 150

🤖 What is Reachy Mini?

Reachy Mini is an open-source desktop humanoid robot by Pollen Robotics:

  • 6-DOF Head - Stewart platform for natural movement
  • Expressive Antennas - Two servos for emotions
  • 360° Body Rotation - Full rotation capability
  • HD Camera - Computer vision ready
  • Microphone & Speaker - Voice interaction
  • Python SDK - High-level control
  • REST API - HTTP control from any language
  • HuggingFace Integration - Deploy AI apps

🛒 Buy Reachy Mini | 📖 Official SDK

💻 Supported Languages

  • Python (SDK + REST API)
  • JavaScript/TypeScript
  • Go
  • Rust
  • Java/Kotlin
  • Swift
  • PHP, Ruby, C#, C++
  • 40+ more via OpenAPI Generator

🎨 What Can You Build?

AI Applications

  • Conversational robots with LLMs
  • Object detection and tracking
  • Face recognition
  • Voice assistants
  • Multimodal AI systems

Interactive Apps

  • Web dashboards
  • Mobile apps (iOS/Android)
  • Desktop applications
  • Home automation
  • Monitoring systems

Research & Education

  • Human-robot interaction
  • Robotics education
  • AI experimentation
  • STEM demonstrations

🛠️ Development

Prerequisites

  • Python 3.12+
  • Claude.ai account
  • (Optional) OpenAPI Generator

Build Instructions

# Clone repository
git clone https://github.com/jjmartres/reachy-mini-sdk-skill.git
cd reachy-mini-sdk-skill

# Initialize environment
make install

# Validate
make validate

# Package
make package

# Output: dist/reachy-mini-sdk.skill

Repository Structure

reachy-mini-sdk-skill/
├── skill/                    # Skill source
│   ├── SKILL.md              # Main file (240 lines)
│   ├── NETADATA.yaml         # Skill metadata
│   └── references/           # References
├── scripts/                  # Build scripts
│   ├── package_skill.py
│   └── quick_validate.py
├── dist/                     # Build output
│   └── reachy-mini-sdk.skill
├── LICENSE      
├── README.md
└── CONTRIBUTING.md

📊 Optimization

Follows official Anthropic guidelines:

  • ✅ 50% smaller SKILL.md (487 → 240 lines)
  • ✅ "Concise is Key" principle
  • ✅ Progressive disclosure pattern
  • ✅ Proper metadata structure
  • ✅ Token efficient (~2,200 vs ~4,500 tokens)

🤝 Contributing

Contributions welcome! See CONTRIBUTING.md.

Ideas:

  • Add examples in new languages
  • Improve documentation
  • Report bugs
  • Share projects

🙏 Acknowledgments

📞 Support

🔗 Links

📈 Changelog

CHANGELOG

⭐ Star this repo if you find it helpful!

Made with ❤️ by Jean-Jacques Martres