Home Assistant Integration

[ALERTua]

First of all, thank you for creating Mindwtr. I'm really enjoying the local-first approach.
I was wondering if there is already a Home Assistant integration, or if one is planned for the future.
Thanks!

[dongdongbh]

No integration today, and none planned yet — mainly because "Home Assistant integration" can mean a few quite different things and I'd rather build the one you actually want.

Some of it already works without either project shipping anything. The self-hosted Mindwtr Cloud server has a bearer-token REST API — GET/POST /v1/tasks, PATCH /v1/tasks/:id, POST /v1/tasks/:id/complete, GET /v1/search — so an HA rest_command can capture to the Inbox or complete a task right now: https://docs.mindwtr.app/developers/cloud-api. The desktop app has its own local API, but it binds to 127.0.0.1 only, so that one only helps for automations running on the same machine: https://docs.mindwtr.app/power-users/local-api.

So: what would you most want it to do? Capture to the Inbox from automations or voice, show Focus/Next on a dashboard, complete tasks from HA, fire automations when a task changes, or actually back a real todo entity? You know the HA side far better than I do — my read is that the first few might only need a documented rest_command recipe, while todo entities would be a custom integration in its own repo, with real decisions to make about projects, statuses, recurrence and polling. Which is closest to what you had in mind?

[ALERTua]

So the modern state of todo and recurring chores in HA is quite grim. Grocy is the only semi-working thing for chores. And I have found no convenient todo integrations except Google Keep, which is garbage for obvious reasons.

I would like Mindwtr integration to be able to do these actions in HA:

• List todo items for a todo list.
• Add/complete a todo item for a list.
• List chores with their next occurrence datetime to be able to notify about an upcoming chore.
• Complete a chore item occurrence.

This would make going to Mindwtr web/app unneeded, which is one of the goals for HA.

A thread in the Home Assistant Community Forums with this question would reveal many more requests.
But it's up to you. If you aren't a HA user, then having no plans for an integration is fully understandable.

[dongdongbh]

That's a clear enough spec to build against, thanks.

All four map onto Home Assistant's native to-do entity, so the shape is: one Mindwtr project = one todo entity, and chores are just tasks in it that carry a recurrence. No custom card or special chore entity needed.

The recurrence part works out better than I expected. Mindwtr keeps exactly one live occurrence of a repeating task, and completing it advances the dates and creates the next one — POST /v1/tasks/:id/complete already does that server-side. So HA would complete the current item and re-poll; the next occurrence's due date lands on the new task, which is what an "upcoming chore" automation would trigger on.

Where the code goes: a separate repo (mindwtr-home-assistant), installed as a custom integration via HACS, not inside the Mindwtr apps. Config flow takes the self-hosted Cloud URL, a bearer token, and which projects to expose. A first version lists open tasks per project, adds tasks, completes them, exposes due dates, and polls on a coordinator with an immediate refresh after each write.

Two limits worth stating up front. Recurrence rules would still be defined in Mindwtr — HA's to-do item model has no recurrence field, so there's nowhere to put "every two weeks after completion". And GET /v1/tasks filters by status and search but not by project today, so v1 would fetch and filter locally; adding ?projectId= is easy if it turns out to matter.

I can't commit to a timeline — I don't run Home Assistant, so anything I wrote would be untested against a real setup. Please do open the Community Forums thread and link it here. If it surfaces people who'd use it that's the strongest argument for building it, and a contributor or testers would make it happen much faster.

[ALERTua]

https://community.home-assistant.io/t/mindwtr-open-source-local-first-gtd-app-is-there-interest-in-a-home-assistant-integration/1019531