home-assistant-agent 0.12.1

Agent for interacting with Home Assistant REST API

Home Assistant Agent - A2A | AG-UI | MCP

Version: 0.12.1

Overview

Home Assistant Agent MCP Server + A2A Agent

Agent for interacting with Home Assistant REST API

This repository is actively maintained - Contributions are welcome!

MCP

Using as an MCP Server

The MCP Server can be run in two modes: stdio (for local testing) or http (for networked access).

Environment Variables

• HOME_ASSISTANT_URL: The URL of the target service.
• HOME_ASSISTANT_TOKEN: The API token or access token.

Run in stdio mode (default):

export HOME_ASSISTANT_URL="http://localhost:8080"
export HOME_ASSISTANT_TOKEN="your_token"
home-assistant-mcp --transport "stdio"

Run in HTTP mode:

export HOME_ASSISTANT_URL="http://localhost:8080"
export HOME_ASSISTANT_TOKEN="your_token"
home-assistant-mcp --transport "http" --host "0.0.0.0" --port "8000"

A2A Agent

Run A2A Server

export HOME_ASSISTANT_URL="http://localhost:8080"
export HOME_ASSISTANT_TOKEN="your_token"
home-assistant-agent --provider openai --model-id gpt-4o --api-key sk-...

Security & Governance

This project is built on agent-utilities, inheriting enterprise-grade security and governance features.

Authentication & Authorization

Feature	Description
OIDC Token Delegation	RFC 8693 token exchange for user-context propagation from A2A → MCP
Eunomia Policies	Fine-grained, policy-driven tool authorization (none, embedded, remote)
Scoped Credentials	Tools execute with the caller's scoped identity where possible
3LO / OAuth / API Token	Multiple auth strategies with graceful fallback

Eunomia Policy Enforcement

Eunomia provides a policy enforcement point for all tool calls:

• Embedded mode: Load local mcp_policies.json for role-based access, sensitivity gating, and audit logging
• Remote mode: Forward authorization decisions to a central Eunomia policy server for multi-agent governance
• Enable via CLI: --eunomia-type embedded --eunomia-policy-file mcp_policies.json

Runtime Protections

Protection	Description
Tool Guard	Sensitivity detection with human-in-the-loop approval gating
Prompt Injection Defense	Input scanning and repetition/loop guards
Content Filtering	Output schema enforcement and cost budget controls
Stuck Loop Detection	Automatic detection and recovery from agent loops
Context Limit Warnings	Proactive alerts before context window exhaustion

Graph Agent Architecture

The A2A agent uses pydantic-graph orchestration with:

• RouterNode: Lightweight classifier that routes queries to specialized domains
• DomainNode: Focused executor with only relevant tools loaded, preventing tool hallucination
• Approval Gates: Policy-driven approval workflows before sensitive operations
• Usage Guards: Budget and rate limiting enforcement

Production Recommendation: Enable --eunomia-type embedded (or remote) + OIDC delegation + containerized deployment. See agent-utilities documentation for full policy configuration.

Docker

Build

docker build -t home-assistant-agent .

Run MCP Server

docker run -d \
  --name home-assistant-agent \
  -p 8000:8000 \
  -e TRANSPORT=http \
  -e HOME_ASSISTANT_URL="http://your-service:8080" \
  -e HOME_ASSISTANT_TOKEN="your_token" \
  knucklessg1/home-assistant-agent:latest

Deploy with Docker Compose

services:
  home-assistant-agent:
    image: knucklessg1/home-assistant-agent:latest
    environment:
      - HOST=0.0.0.0
      - PORT=8000
      - TRANSPORT=http
      - HOME_ASSISTANT_URL=http://your-service:8080
      - HOME_ASSISTANT_TOKEN=your_token
    ports:
      - 8000:8000

Configure mcp.json for AI Integration (e.g. Claude Desktop)

{
  "mcpServers": {
    "home-assistant": {
      "command": "uv",
      "args": [
        "run",
        "--with",
        "home-assistant-agent",
        "home-assistant-mcp"
      ],
      "env": {
        "HOME_ASSISTANT_URL": "http://your-service:8080",
        "HOME_ASSISTANT_TOKEN": "your_token"
      }
    }
  }
}

Install Python Package

python -m pip install home-assistant-agent

uv pip install home-assistant-agent

Repository Owners

MCP Configuration Examples

stdio (recommended for local development)

{
  "mcpServers": {
    "home-assistant": {
      "command": ".venv/bin/home-assistant-mcp",
      "args": [],
      "env": {
        "HOME_ASSISTANT_URL": "",
        "HOME_ASSISTANT_TOKEN": ""
}
    }
  }
}

Streamable HTTP (recommended for production)

{
  "mcpServers": {
    "home-assistant": {
      "url": "http://localhost:8080/home-assistant-mcp/mcp"
    }
  }
}

Available MCP Tools

This server utilizes dynamic Action-Routed tools to optimize token overhead and maximize IDE compatibility.

Tool Name	Description
ha_calendar	Consolidated Action-Routed tool for calendar. Methods: list_calendars, get_calendar_events
ha_config	Consolidated Action-Routed tool for config. Methods: status, config, components, check_config
ha_entities	Consolidated Action-Routed tool for entities. Methods: get_entity_registry_display, extract_from_target, get_triggers_for_target, get_conditions_for_target, get_services_for_target
ha_events	Consolidated Action-Routed tool for events. Methods: list_events, fire_event, subscribe_events
ha_history	Consolidated Action-Routed tool for history. Methods: get_history
ha_logbook	Consolidated Action-Routed tool for logbook. Methods: get_logbook, get_error_log
ha_panels	Consolidated Action-Routed tool for panels. Methods: get_panels
ha_services	Consolidated Action-Routed tool for services. Methods: list_services, call_service
ha_states	Consolidated Action-Routed tool for states. Methods: list_states, get_state, update_state, delete_state
ha_system	Consolidated Action-Routed tool for system. Methods: render_template, ping, handle_intent, validate_config
ha_voice	Consolidated Action-Routed tool for voice. Methods: list_exposed_entities, expose_entities