AI-guided Home Assistant setup — automation templates, mobile-first React dashboard, with Claude Code skills for configuration management

Home Assistant Claude Kit Github Repository

For the last couple of weeks I’ve spent some time completely overhauling my HA setup using Claude Code to create over 100 automations and a completely custom dashboard centered around how we use our home.

I’m a software developer and in particular one of those who didn’t write a single line of code since December, although I’ve written more code this year than in the last 5 years. So I applied the same methods that I use in my job with HA.

Why This Exists

I’ve been looking for a long time for a truly customizable dashboard that would be based on how you actually use the house, that is able to mix seamlessly entities from different integrations and more importantly have the interface fully customized to control the home through your own custom automations, rather than just be a collection of device cards. And the reason something like that isn’t generally available is because it requires a lot of custom development for each home and as flexible, user friendly and complex a dashboard admin is it will never be able to reach the level of customization of pure code custom built. And that takes a lot of time and effort.

I don’t care about having a boiler card on my dashboard and 7 TRV controls. I care about having a fully automated multi-zone heating system that coordinates the boiler with the TRVs, with day/night schedule and manual overrides. And this is one of the things I’ve built with this system. There are no boiler or TRV controls in my dashboard, Just temperature and schedule controls in the system settings view and on the Climate view you only have status and manual overrides.

Same with the EV charger, dead simple controls: Off, Solar, Fast, Manual. Don’t think they need much explaining other than the Solar mode only starts the car charging when there is enough solar production and adjusts it as production varies with adaptable hysteresis and tolerances for sunny/cloudy days.

And AI changed all that. There is no visual editor for the dashboard; it’s custom React code. There is no UI for editing automations; everything lives in YAML files managed through Claude Code. You describe what you want in plain language, and the AI writes the code, validates it, deploys it, and documents it. Yes, there are risks in AI-generated code, but they are mitigated by validation at every step (pre-push hooks, TypeScript checks, entity reference validation), a structured workflow that catches issues early, and tight integration with HA through CLI tools that eliminate manual copy-paste and command-running.

Design Principles

  • Use-case driven, not device-driven. The dashboard and automations are organized around how you use your home – climate zones, activity modes, energy management – not around individual devices or integrations.
  • Nothing hardcoded. Every threshold, timeout, temperature, brightness, and duration is an input_* helper, visible and editable in the Settings view. Automations read these at runtime – no magic numbers, everything tweakable without touching YAML.
  • Self-documenting. docs/system-*.md describes what IS (hardware, entities, design decisions – generated during setup, populated as you build). docs/solutions/ captures debugging lessons (34 docs and growing). Both are updated as part of every change, so documentation never drifts from reality.
  • Mobile-first. The dashboard is designed for phones. Every interaction is touch-optimized. When a design compromise was needed, mobile experience was favored over desktop.
  • Validate before deploy. Hooks run on every file edit. make push validates before syncing. Broken configs never reach HA.
  • Event-driven, not polling. Automations use state triggers with for: durations, not timers or delays. The system reacts to changes rather than checking on intervals.

Before you start

It is worth noting that the dashboard is just the tip of the iceberg. When I stated this I barely had 15 automations in my system, now there are 107 and 12 scripts. That’s where I spent most of my time, designing and fine tuning the automations, although Claude has written all the code.
And although I am sure all of you know AI is not a magic bullet, I cannot stress enough how important it is to use proper workflows for development. I strongly recommend the Compound Engineering workflow that I fully documented in the prompting guide. Deepening and reviewing the plans (at least 2-3 times depending on the complexity) before any actual code is written is the single most important step you should never skip for any complex work. Also after the implementation reviewing the code (several times) and compounding is almost just as important.

Although it started as a personal project for my own usage, for the last couple of days I have extracted all the reusable automations, the dashboard components, added an interactive HA entities discovery and setup skill for Claude, so that the kit can be used on any HA install. You can use it as a starting base to create your own personal, completely customized HA automations and dashboard centered around those automations with Claude Code without touching a single line of YAML or TypeScript in a fraction of the time it would take to do it yourself.

The project is developed specifically with and for Claude Code, but it was already ported to OpenClaw and successfully used with minimax using the Claude plugin porting scripts in the Compound Engineering Claude plugin repository that you can find in my project’s acknowledgments.

Home Assistant Claude Kit Github Repository

1 Like