Indigo FAQs

Question 1

What core smart home functions can I control using natural language with Indigo?

Accepted Answer

You can semantically search for any Indigo entity (devices, variables, action groups), directly control devices (turn on/off, set brightness), update variable values, execute action groups (scenes), and perform AI-powered analysis of historical device data and trends.

Question 2

What are the security and privacy considerations when using Indigo?

Accepted Answer

For semantic search, device names, states, and descriptions are sent to OpenAI for embedding generation. However, only minimal, non-sensitive data is processed, and all vector embeddings are stored locally. By default, the HTTP server binds to localhost, ensuring it's not exposed to the internet unless explicitly configured otherwise.

Question 3

Are there any specific requirements or integrations needed to use Indigo?

Accepted Answer

Yes, Indigo requires an OpenAI API Key for semantic search capabilities and Indigo Domotics 2024.2 or later with Python 3.11+. Optional integrations include LangSmith for advanced debugging and InfluxDB for historical data analysis.

Question 4

What is Indigo and how does it enhance my smart home?

Accepted Answer

Indigo is an MCP server plugin for Indigo Domotics that bridges AI assistants like Claude with your home automation system. It enables natural language control, semantic search of devices/variables/actions, and AI-powered historical data analysis, making your smart home more intuitive.

Question 5

Which AI assistants are compatible with the Indigo plugin?

Accepted Answer

The Indigo plugin is fully tested and supported with Claude Desktop. It is designed to work with any MCP-compatible AI client using its HTTP transport endpoint, offering broad compatibility.

Indigo

关于

主要功能

使用案例