How it works

Push it in, and then mostly forget about it.

There is no app to configure for an hour, no hub to plug into your router, and nothing to calibrate. Three steps, and then it is just part of the pot.

Push it into the soil.

The stake goes in next to the plant, deep enough that the probe sits where the roots are rather than in the dry crust on top. That is the entire installation. It takes about four seconds and you will probably do it wrong the first time, which is fine.

A green Sprooty frog on a blue stake pushed into the soil of a potted plant on a sunny windowsill.

deep enough to reach the roots

It watches, day and night.

Three readings, taken regularly and sent over your home Wi-Fi. Individually they are just numbers. Together they explain almost everything that goes wrong with a houseplant.

An orange Sprooty frog sitting in a terracotta pot of dry ornamental grass. — It stays put. You get on with your life.

How wet the soil is

The one that matters most, and the one people are worst at guessing.

How warm it is

A windowsill in February is a different room from the rest of your flat.

How much light it gets

Most "why is it dying" answers are really "it is in the wrong corner".

Your phone says what to do.

Not a graph for you to interpret — a sentence. “The fiddle fig has been dry for six days.” And if you want to know why that keeps happening, you can ask, and the answer will be about your plant in your room rather than a search result.

9:41

Ask Sprooty

Why is my fiddle fig dropping leaves?

Its soil has been under 20% for six days, and the room dropped to 16°C on Tuesday. Fiddle figs drop leaves when they get cold and dry at once.

Water it properly, then move it away from the window at night.

Ask about a plant…

9:41

Your plants

One of them needs you today

Monstera

Happy

Living room

Fiddle fig

Thirsty

Bedroom

Peace lily

A bit wet

Kitchen

PlantsAskRoomsYou

For the curious

Under the hood

Skip this bit unless you build things. It is here because people ask, and because we would rather be specific than vague.

The frog: An ESP32 with a capacitive soil-moisture probe plus light and temperature sensors, sleeping between readings and waking to post them over Wi-Fi.
The backend: FastAPI on Python 3.12. Postgres holds users, plants and devices; InfluxDB holds the sensor time series, because that is what it is for.
The assistant: An LLM given your plant, your room and your recent readings as context, so the advice is grounded in your data rather than generic.
The app: React Native and Expo, one codebase for iOS and Android.

Want one when they exist?

Join the list