ITSpecialist111/ai_automation_suggester

This custom Home Assistant integration automatically scans your entities, detects new devices, and uses AI (via cloud and local APIs) to suggest tailored automations. It supports multiple AI providers, including OpenAI, Anthropic, Google, Groq, Ollama and more! The integration provides automation suggestions via HASS notifications

What it solves

This project provides an AI-powered assistant for Home Assistant that helps users overcome "automation writer's block" and underutilized smart home hardware. It analyzes a user's specific smart home environment—including devices, areas, and existing automations—to suggest intelligent, actionable YAML configurations for new automations, reducing the effort and guesswork involved in designing complex smart home behaviors.

How it works

The integration follows a five-step process to generate suggestions:

  1. Snapshot: It collects data on entities, devices, areas, and existing automations, with configurable filters and limits.
  2. Prompt Building: This data is embedded into a detailed system prompt, which can be further customized by the user to steer the AI toward specific goals.
  3. Provider Call: The prompt is sent to a configured AI provider (supporting a wide range of cloud services like OpenAI, Anthropic, Google, and local options like Ollama or LocalAI).
  4. Parsing: The integration processes the AI's response, preferring structured JSON or falling back to fenced YAML parsing and validation.
  5. Surface: Suggestions are delivered as Home Assistant persistent notifications, sensor attributes, and a stored history for use in dashboards.

Who it’s for

Home Assistant users who have complex smart home setups and want to discover new automation possibilities or need help translating their ideas into functional YAML code.

Highlights

  • Broad Model Support: Compatible with over 100 backends via providers like LiteLLM, OpenRouter, and native integrations for OpenAI, Anthropic, and Google.
  • Context-Aware: Uses actual device and area data from the user's home for realistic recommendations.
  • Customizable Control: Offers domain filters, entity limits, and custom system prompts to refine the AI's output.
  • Safe Implementation: Suggestions are presented for review and must be manually implemented by the user.
  • Integrated Workflow: Provides built-in example automations for weekly reviews or triggering suggestions when new entities are added.

Related

  • Project
    ThomasVitale/llm-apps-java-spring-ai

    A collection of sample Java applications and patterns demonstrating how to build Generative AI and LLM-powered apps using the Spring AI framework.

  • Project
    line/line-bot-mcp-server

    An MCP server that integrates the LINE Messaging API, allowing AI agents to send messages and manage LINE Official Accounts.

  • Project
    Jia-Ethan/claude-keysmith

    A tool for adding revocable custom instructions to Claude Code, allowing users to preview changes before applying them to new AI conversations.

  • Project
    assistant-ui/tool-ui

    A collection of copy-paste React components that transform raw AI tool-call JSON into interactive UI elements like charts, forms, and maps.