komal-skynet/claude-skill-homeassistant
Overview
This skill provides expert Home Assistant configuration management and rapid deployment workflows to speed up testing and production rollouts. It combines SSH and hass-cli remote access, git and scp deployment patterns, automation verification, log analysis, and full Lovelace dashboard management optimized for tablets. Practical templates, card patterns, and debugging strategies are included to reduce downtime and speed iteration.
How this skill works
It inspects and manages Home Assistant configuration by using hass-cli for REST interactions and SSH for direct host operations (ha core commands). Deployments use two workflows: git for tracked, production-ready changes and scp for fast iterative testing. The skill verifies automations, analyzes logs for errors, decides between reload vs restart, and edits Lovelace JSON for responsive, tablet-optimized UIs.
When to use it
- Use scp for rapid UI/dashboard iteration and experimental changes.
- Use git for final, version-controlled changes and rollbacks.
- Run hass-cli for state inspection, service calls, and quick checks without SSH.
- Use SSH when you must run ha core checks, view host logs, or perform restarts.
- Optimize reload vs restart when you need minimal disruption.
Best practices
- Always run ha core check via SSH before committing or pulling changes.
- Prefer service reloads (automation/script/scene/template) when possible to avoid full restarts.
- Use scp + browser hard-refresh for instant Lovelace feedback; commit only stable dashboard files.
- Wrap template numeric comparisons with |int or |float to avoid type errors.
- Tail logs and grep for automation names after manual trigger to confirm outcomes.
Example use cases
- Iterate a tablet control panel: edit .storage/lovelace.control_center locally, scp to instance, refresh browser, repeat until good, then git commit.
- Deploy automations: commit to repo, ssh pull on instance, run ha core check, reload automations, manually trigger and inspect logs.
- Debug failing card: check browser console, validate JSON, verify custom card installed via HACS, confirm entities with hass-cli.
- Recover after config change: ssh run ha core check, inspect logs for errors, revert via git or patch file and reload/restart as needed.
FAQ
Restart only for platform additions, core configuration changes, MQTT/platform sensors, or when ha core check reports issues that reload cannot fix.
How do I quickly verify an automation worked?
Manually trigger it with hass-cli service call automation.trigger, wait a few seconds, then grep ha core logs for the automation name and check the expected entity states with hass-cli state get.