Hi,

I love jupyter for creating, maintaining and debugging my HA scripts and templated sensors :)

I got some templated sensors that just get too verbose in jinja2.

So I created a jupyter notebook and proved I could create the same output including updating the already defined sensor... Perfect!

And I found you can use nbconvert to convert the notebook to Python - so far so good..

I could easily see my script nbconvert and upload to HA when ever I save the notebook or similar.

But I had to inject a

@state_trigger("sensor.energy_production_remaining or sensor.energi_data_service_buy or sensor.sonnen_battery_level_kwh or sensor.sonnen_hours_to_charge_window_end", state_check_now=True)
def update_sonnen_needed_charge_hours():

between from datetime import datetime and my actually code.

This won't be the last time I update this script and don't like the idea of me having to copy the changes over manually..

How do everyone else keep your Notebook testbeds synchronized with code in production?