Overview
What if your AI assistant could do more than answer questions — what if it could control your entire smart home? This project demonstrates exactly that by integrating Shelly Cloud APIs directly into Claude, turning the AI into a powerful home automation hub.
🧠 The Goal
The aim was to link daily-use AI assistants (Claude, Gemini, ChatGPT, or Copilot) with Shelly’s smart ecosystem — without relying on third-party bridges. The solution: a custom-built MCP (Model Context Protocol) server running locally, designed to connect Shelly Cloud APIs straight to the AI.
🔧 How It Works
Local MCP Server: Acts as a translator between the AI and the Shelly Cloud APIs.
Direct API Calls: The AI can fetch device states, trigger actions, or check automation logic.
Natural Commands: Users simply ask Claude things like “Turn off the living room lights” or “Check if the garage door is closed,” and the AI responds instantly.
💡 What’s Possible
This setup means:
All Shelly devices (lights, plugs, relays, sensors) are instantly controllable via AI.
No app-switching — everything happens through conversational requests.
It’s scalable to other AI assistants (Gemini, ChatGPT, Copilot, etc.) with the same MCP server.
🔍 Why It Matters
Natural control: Talking to an AI to manage your home feels seamless and intuitive.
Flexible integrations: No lock-in — this works with multiple assistants and platforms.
Future-ready: Opens the door for more complex automations, routines, and data-driven triggers.
🌟 The Outcome
Now, the AI isn’t just conversational — it’s actionable. Through Claude, every Shelly device in the home can be accessed, switched, or monitored. It’s the next step in making home automation invisible, intuitive, and intelligent.
