hass-mcp

Local
Community
272
GitHub Repo

Overview

The hass-mcp server is a Model Context Protocol (MCP) server that allows AI assistants and agents to interact directly with Home Assistant (HASS), the open-source home automation platform. It enables AI-driven workflows to observe, control, and reason about smart home entities and states — turning a Home Assistant instance into a programmable, conversational environment.

This server is ideal for smart home automation, contextual assistance, energy monitoring, and ambient intelligence use cases where AI needs real-time awareness of a physical environment.

Transport

stdio

Tools

  • get_version
  • get_entity
  • entity_action
  • list_entities
  • search_entities_tool
  • domain_summary_tool
  • list_automations
  • call_service_tool
  • restart_ha
  • get_history
  • get_error_log

Key Capabilities

  • Real-time state awareness — Access live state information for sensors, devices, and automations.
  • Device and service control — Enable AI assistants to control lights, climate, switches, and other smart devices.
  • Contextual automation — Combine AI reasoning with Home Assistant’s existing automations and scenes.
  • Home topology understanding — Reason about areas, devices, and entity relationships.
  • Privacy-preserving operation — Run locally alongside Home Assistant without requiring cloud services.

How It Works

The hass-mcp server runs as a local MCP service that connects to a Home Assistant instance using a long-lived access token. AI clients communicate with the server over the MCP protocol to retrieve state, inspect configuration, or request actions as part of broader reasoning workflows.

The server mediates all interaction with Home Assistant’s APIs, ensuring that access respects configured permissions and scopes. Information about entities, devices, and areas is returned in structured formats that AI assistants can reason over and combine with context from other MCP servers.

By exposing Home Assistant through MCP, the server enables AI-driven smart home experiences — such as conversational control, context-aware suggestions, and environment-aware automation — without embedding Home Assistant–specific logic into the assistant itself.