Provides control over Alexa-connected smart home devices through the Alexa API, enabling voice announcements, music control, smart lighting automation, sensor monitoring, and volume management across Echo devices and connected smart home accessories.
Alexa Home Automation MCP Server ๐ ๐
A powerful Model Context Protocol (MCP) server that provides AI agents (like Claude Desktop) with full control over your Alexa-connected smart home.
This server acts as a bridge between the Alexa API and the MCP ecosystem, enabling capabilities like voice announcements, music control, lighting automation, and sensor monitoring.
โจ Features
๐ข Voice Announcements: Send customized voice messages to any Alexa device with smart night-time suppression.
๐ต Music Control: Get real-time status of current tracks and playback on your Echo devices.
๐ก Smart Lighting: Full control over power, brightness, and colors for all discovered smart lights.
๐ก๏ธ Sensor Integration: Access temperature, light, and motion sensor data from your Alexa devices.
๐ Volume Management: Precise control over volume levels across your entire device fleet.
๐ค Self-Documenting: Clean, descriptive API handlers designed for LLM consumption.
๐ Getting Started
1. Requirements
pnpm installed.
An Amazon account with connected Alexa devices.
2. Authentication (Automated Cookie Collection)
We've automated the tedious process of collecting Amazon cookies.
Install dependencies:
pnpm installRun the cookie automation script:
pnpm run get-cookiesA local proxy will start. Complete the login in the browser window that appears.
The script will automatically generate your
.envfile with the requiredUBID_MAINandAT_MAINtokens.
3. Local Development
Start the server in Node.js mode:
The server will be available at http://localhost:3001.
๐ ๏ธ MCP Configuration
Add this to your Claude Desktop configuration (claude_desktop_config.json):
๐๏ธ Architecture
This project is built using:
Hono: A fast, lightweight web framework.
TypeScript: For robust, type-safe development.
MCP SDK: To enable seamless integration with AI agents.
Cloudflare Workers Ready: Can be deployed to Cloudflare or any Node.js environment via Docker.
๐ License
MIT ยฉ sijan2