Atlas / MCP servers / jango-blockchained / Advanced Home Assistant

Advanced Home AssistantCAUTION

Name: Advanced Home Assistant
Author: jango-blockchained

mcp/jango-blockchained/advanced-home-assistant

MCP server by jango-blockchained · home assistant · llm · mcp server · Audited 2026-10-08

MCP server that gives AI assistants control over your Home Assistant. Including 50+ tools over 3 transports. Batteries included 🔋

Verdict

CAUTION

Grade

Trust score

89 /100

Exposed tools

91 68r · 20w · 3d

Transport

—

License

Apache-2.0

Stars

Overview

From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.

[](https://smithery.ai/server/@jango-blockchained/advanced-homeassistant-mcp) [](https://github.com/jango-blockchained/advanced-homeassistant-mcp/releases) [](https://www.npmjs.com/package/@jango-blockchained/homeassistant-mcp) [](https://github.com/jango-blockchained/advanced-homeassistant-mcp/pkgs/container/advanced-homeassistant-mcp) [](https://github.com/jango-blockchained/advanced-homeassistant-mcp/actions) [](https://bun.sh) [](https://nodejs.org) [](LICENSE)

Talk to your house. Your AI assistant (Claude, GPT, Cursor, Copilot) — connected to Home Assistant through the Model Context Protocol. 50+ tools, three transports, one bunx command.

# ⚡ Fastest way to try it
bunx github:jango-blockchained/advanced-homeassistant-mcp

🚀 Quick Start

You need a Home Assistant instance and a long-lived access token (create one here).

Claude Desktop

Add to claude_desktop_config.json:

{
"mcpServers": {
"homeassistant-mcp": {
"command": "bunx",
"args": ["github:jango-blockchained/advanced-homeassistant-mcp"],
"env": {
"HASS_HOST": "http://your-ha-instance:8123",
"HASS_TOKEN": "your_long_lived_access_token"
}
}
}
}

**R

Read from source at commit d2f68b18429dOBSERVED · 2026-10-08

Connect

Built from this server's own package name, version and transport as found in its source — not copied from anyone's documentation, so it cannot drift against a page we do not control. Replace the environment placeholders with a token scoped to the least it needs.

claude-code

claude mcp add homeassistant-mcp --env ANTHROPIC_API_KEY=${ANTHROPIC_API_KEY} --env COOKIE_SECRET=${COOKIE_SECRET} --env CORS_CREDENTIALS=${CORS_CREDENTIALS} --env HASS_TOKEN=${HASS_TOKEN} -- npx -y @jango-blockchained/[email protected]

claude-desktop

{
  "mcpServers": {
    "homeassistant-mcp": {
      "command": "npx",
      "args": [
        "-y",
        "@jango-blockchained/[email protected]"
      ],
      "env": {
        "ANTHROPIC_API_KEY": "${ANTHROPIC_API_KEY}",
        "COOKIE_SECRET": "${COOKIE_SECRET}",
        "CORS_CREDENTIALS": "${CORS_CREDENTIALS}",
        "HASS_TOKEN": "${HASS_TOKEN}"
      }
    }
  }
}

Exposed tools (91)

68 read · 20 write · 3 destructive. Blast radius: 3 tools can delete or overwrite — an agent that can be talked into calling a tool can be talked into calling this one.

Tool	Risk	Description
`Automations`	read	List of all configured automations
`Covers`	read	All cover entities (blinds, curtains, garage doors)
`Fans`	read	All fan entities and their states
`Health`	read	Health check endpoints for monitoring server status
`Locks`	read	All lock entities and their states
`MCP`	read	Model Context Protocol endpoints for tool execution
`SSE`	read	Server-Sent Events endpoints for real-time updates
`Scenes`	read	List of all configured scenes
`Sensors`	read	All sensor entities (temperature, humidity, etc.)
`Switches`	read	All switch entities
`Tools`	read	Tool management endpoints for listing and configuring tools
`action`	read	What to do with the lights (e.g.,
`activity`	read	What you want to do (e.g.,
`addon`	write	List Home Assistant add-ons or get info on a specific add-on
`addon_modify`	destructive	Manage Home Assistant add-ons (install, uninstall, start, stop, restart)
`alarms`	read	List all alarm panels or get the state of a specific alarm panel.
`alarms_activate`	read	Arm in various modes (home, away, night, vacation, custom bypass), disarm, or trigger an alarm. Security-sensitive operations.
`animation_control`	read	List currently running background light animations.
`animation_control_activate`	write	Start, stop, or stop-all background light animations (endless loops).
`automation`	read	List Home Assistant automations and their current state.
`automation_activate`	read	Fire a Home Assistant automation
`automation_config_modify`	write	Manage Home Assistant automations
`automation_modify`	write	Toggle (enable/disable) a Home Assistant automation. Does not fire the automation
`claude`	read	Claude-specific prompt template for home automation control
`climate`	read	List all climate devices (thermostats, AC) or get the state of a specific one.
`climate_activate`	write	Set HVAC mode, target temperature, or fan mode on climate devices.
`climate_comfort`	read	Optimize climate control for comfort and efficiency
`control`	read	Control Home Assistant entities
`control_activate`	read	Control Home Assistant entities (lights, climate, etc.)
`control_lights`	read	Help with controlling lights in a room or area
`covers`	read	List all covers (blinds, curtains, garage doors, shades) or get the state of one.
`covers_activate`	write	Open, close, stop, toggle, or set the position/tilt of a cover.
`custom`	read	Customizable prompt template for home automation control
`dashboard`	read	List Home Assistant Lovelace dashboards or read a dashboard
`dashboard_modify`	write	Replace a Home Assistant Lovelace dashboard
`device_name`	read	The name or type of device having issues
`energy_saving`	read	Suggestions for energy-saving automations
`fans`	read	List all fans or get the state of a specific fan.
`fans_activate`	write	Control fans: turn on/off/toggle, set speed percentage, preset mode, oscillation, direction.
`get_entity_state`	read	Get the current state of any Home Assistant entity. Works with all entity types: sensors, binary_sensors, switches, lights, climate, covers, etc. Returns state value, timestamps, and optionally all attributes.
`get_history`	read	Retrieve historical data for entities
`get_sse_stats`	read	Get statistics about SSE connections
`gpt4`	read	GPT-4 specific prompt template for precise home automation control
`history`	read	Get entity history
`light_animation_activate`	write	Execute custom light animation sequences (e.g., police lights, alerts) on any RGB light.
`light_scenario_activate`	write	Apply ambient lighting moods (Chill, Nightly, Focus, etc.) to a specific area or light.
`light_showcase_activate`	write	Run a choreographed light showcase that calls various moods and custom palettes to demonstrate system capabilities.
`lights`	read	List all lights or get the state of a specific light.
`lights_activate`	read	Turn lights on (with optional brightness/color/effect) or off.
`list_devices`	read	List all Home Assistant devices
`locks`	read	List all locks or get the state of a specific lock.
`locks_activate`	read	Lock, unlock, or open (unlatch) physical locks. Unlocking is a security-sensitive operation.
`maintenance`	read	Analyze Home Assistant health: find orphaned/unavailable devices, analyze light usage and energy consumption, run device health checks. All actions are read-only — no entities are removed or modified.
`media_control`	read	Control media players and entertainment systems
`media_players`	read	List all media players or get the state of a specific media player.
`media_players_activate`	read	Control media players: playback, volume, source/sound mode selection, play media on TVs and speakers.
`morning_routine`	write	Create a morning routine automation
`notify`	write	Send notification
`notify_activate`	write	Send notifications through Home Assistant
`package`	read	List HACS packages and custom components by category
`package_modify`	destructive	Manage HACS packages (install, uninstall, update)
`preferred_temp`	read	Your preferred temperature (e.g.,
`render_template`	read	Evaluate a Home Assistant Jinja2 template. Use this to query complex state information, perform calculations, format data, or test template expressions. The template is rendered server-side by Home Assistant and the result is returned as text.
`room`	read	The room or area name (e.g.,
`scene`	read	List Home Assistant scenes.
`scene_activate`	read	Activate a Home Assistant scene
`scene_creation`	write	Create a scene for a specific activity or mood
`scene_name`	read	Name for the scene (e.g.,
`search_entities`	read	Search Home Assistant entities with powerful filtering. Supports domain, device_class, state (exact or comparison like
`security_setup`	write	Set up security and monitoring
`smart_scenarios`	read	Detect common smart home scenarios (nobody home, window/heating conflicts, energy-saving opportunities). Read-only analysis.
`smart_scenarios_activate`	write	Apply smart home scenarios: turn off lights and adjust climate when nobody is home, or turn off heating when windows are open. Actuates devices and sends notifications.
`stream_generator`	read	Generate a stream of data with configurable delay and count
`subscribe_events`	read	Subscribe to Home Assistant events via SSE
`switches`	read	List all switches or get the state of a specific switch (smart plugs, relays, virtual switches).
`switches_activate`	read	Turn switches on, off, or toggle.
`system_info`	read	Get basic information about this MCP server including version, transport, and connection status
`test_tool`	read	A test tool
`text_to_speech`	read	Generate and play text-to-speech audio via Home Assistant. Provides voice feedback for commands with support for multiple languages and TTS providers.
`todo`	read	List Home Assistant to-do lists or get items from a specific list.
`todo_modify`	destructive	Add, update, or remove items in a Home Assistant to-do list.
`trace`	read	Access automation and script execution traces. List recent traces, get detailed trace data, or list trace contexts. Useful for debugging automation issues.
`troubleshoot_device`	read	Help troubleshoot a device that
`vacuum_schedule`	write	Set up robot vacuum cleaning schedule
`vacuums`	read	List all robot vacuums or get the state of a specific vacuum.
`vacuums_activate`	write	Control robot vacuums: start/pause/stop, return to dock, spot cleaning, locate, fan speed, vendor-specific commands.
`validation_demo`	read	Demonstrates parameter validation with Zod schemas
`voice_command_ai_parser_activate`	read	Parse voice commands using AI (Claude). Makes an outbound API call to api.anthropic.com (consumes API credits, sends transcript to a third party). Falls back gracefully if AI not available.
`voice_command_executor_activate`	write	Execute parsed voice commands through Home Assistant. Maps intents to service calls and controls devices.
`voice_command_parser`	read	Parse natural language voice transcriptions into structured Home Assistant commands. Extracts intent, entities, and parameters from voice input.
`voice_control_setup`	read	Recommendations for voice control setup

Trust audit

CAUTIONgrade B · trust 89/100 Install with care. The audit found things worth knowing before you trust its output.

Layer	What it checks	Result
L0	Provenance & inventory	PASS
L1	Static analysis of the code	WARN
L2	Instruction surface (what it tells the agent)	PASS
L3	Class-specific surface	WARN
L4	Behavioural (sandbox)	SKIPPED

What the source does

Filesystem: declared (2 observation(s))
Network: declared (2 observation(s))
Shell: declared (1 observation(s))
Dependencies: not all pinned
Secrets in source: none-found

Findings (19)

MEDIUMInformation disclosure · disclose.log_secret · CWE-209, CWE-532

src/hass/index.ts:22

logger.debug(`Token loaded: yes (${this.token.length} chars)`);

MEDIUMFilesystem / path · mcp.destructive_tools · CWE-22, CWE-59

addon_modify, package_modify, todo_modify

Why it matters. 3 tool(s) can delete or overwrite

Fix. prefer a read-only mode or scoped tokens; the page states the blast radius

LOWFilesystem / path · fs.traversal · CWE-22, CWE-59

__tests__/ai/endpoints/ai-router.test.ts:4

import type { AIResponse, AIError } from "../../../src/ai/types/index.js";

LOWFilesystem / path · fs.traversal · CWE-22, CWE-59

__tests__/ai/endpoints/ai-router.test.ts:23

void mock.module("../../../src/ai/nlp/processor.js", () => ({

LOWFilesystem / path · fs.traversal · CWE-22, CWE-59

__tests__/ai/endpoints/ai-router.test.ts:38

import router from "../../../src/ai/endpoints/ai-router.js";

LOWFilesystem / path · fs.traversal · CWE-22, CWE-59

__tests__/ai/nlp/intent-classifier.test.ts:2

import { IntentClassifier } from '../../../src/ai/nlp/intent-classifier.js';

LOWFilesystem / path · fs.traversal · CWE-22, CWE-59

__tests__/api/index.test.ts:6

import type { Entity } from '../../src/types/hass.js';