Pick the thing you can see, and read what causes it
Live · This works today.
Open the plant, tap Match the symptoms yourself, and pick the thing you can see. The list is what people actually report on that crop, in ordinary words — holes chewed in the leaves, a dark sunken patch on the fruit, wilts in the afternoon and recovers overnight. Choose one and you get what causes it, in the words of the university or extension service that wrote it, with their name and a link so you can read the rest. It costs you nothing and uses none of your AI questions — it is a lookup, not a conversation, so you can open it as often as you like. If we hold nothing for that crop, it says so rather than guessing, and offers the problems that are common on most plants instead. And where the cause is a growing problem rather than an illness — blossom end rot on a pepper, say — it tells you that first, so you do not go looking for a spray you do not need. On a crop with a long list there is a box to narrow it down: type leaves and only the leaf ones stay.
Where it is
Open any plant in your garden. On its Story tab, near the foot, there is a quiet line:
Match the symptoms yourself Free · every line quoted with its source
Just above it, Saw a problem › asks first — what you see, and since when — and fills in what the app already knows about the plant; its Match it myself brings you to the same place. This line said Something looks wrong? followed by a paragraph until 1 September 2026 — same door, same place, fewer words, and the two things worth knowing before you tap are now on the face of it: this one costs you nothing, and everything it tells you is quoted from somebody who put their name to it. It is deliberately not a button competing with the ones you press every day — most days nothing is wrong, and a big red DIAGNOSE on every plant would make the garden feel like a hospital.
What you get
A list of things people actually see on that crop, written the way you would say them out loud. Not Xanthomonas campestris — small dark spots with a yellow ring, on older leaves first.
Pick one and you get:
- what causes it, in the source’s own words
- who wrote that, by name
- a link to the rest of what they said
On a crop with a lot of entries — a tomato has sixteen, a broccoli twenty-eight — there is a box above the list. Type leaves and only the leaf ones stay.
What it will not do
It will not guess. If we hold nothing written up for that crop, it tells you plainly and shows the problems common to most plants instead. That is a smaller answer than you wanted, and it is a true one.
It will not tell you which cause is yours. Four things cause a hole in a leaf. This shows you all four and who says so; deciding between them is a job for the chat, where you can send a photo and talk about your particular plant.
When it is not an illness, it says so first
Blossom end rot is the clearest case. It looks alarming — a dark, sunken, leathery patch on the bottom of the fruit — and it is not an infection at all. Nothing has attacked your pepper. It is the plant reacting to how evenly it has been watered.
So the guide says that before it says anything else:
Not an illness — nothing has infected the plant. This is it reacting to how it has been grown, which means it is yours to change.
Twenty-nine entries across the guide are growing problems rather than illnesses, and each one now carries that line. That is three of them — blossom end rot, sunscald and leaf roll — counted once for every plant and symptom they turn up under, which is how many times you could actually meet one.
A cause we hold no write-up for says nothing at all. It does not get called a disease and it does not get called a disorder — because an unknown shown as an illness is exactly how a plant doing something perfectly normal gets reported as sick.
Why it is a list and not a chat
Because a list costs nothing.
The chat is the better tool when you can describe what you are seeing, or photograph it, or when the answer depends on what you did to that plant in June. But it uses an AI question, it needs you to find the words, and it takes a moment. Some days you just want to know what makes a dark patch on a pepper, and a list of the four known answers is faster and cheaper than a conversation.
Both doors lead to the same place. None of these? at the bottom of the list opens the chat with the plant already attached.
Common questions
- It said it thought it could see something on my plant. What is that?
When the app looks at a photo from your garden it writes down what it thinks it can see — first blooms, pods filling, lower leaves going yellow. From 1 September 2026 that turns up on the plant's own page, in the block under its photo, with the reason it gave.
Two buttons: Yes, I can see it and No, not this one.
Saying no is worth as much as saying yes. A yes lets the app use what it saw. A no means it will not come back, and it tells us we misread a picture — which is the most useful thing we can learn about how well this works.
Nothing is used until you answer. ⚠ Until then it is only what we thought, and the app will not act on it, mention it in your week, or count it as having happened. That is on purpose: nothing about your garden is true because a computer said so.
It asks once per thing, however many photos it saw it in. If the app spots the same thing in three pictures of one plant, that is one question, and your answer settles all three. ⚠ Before 1 September 2026 it was one card per picture, which meant answering the same question several times and watching cards seem to come back.
And you do not have to answer. Leaving it alone changes nothing at all.
- Does this use up my AI questions?
No. This is a lookup in records we already hold, not a conversation with an AI, so you can open it as many times as you like and nothing is counted against you. The chat is for when you want to describe something in your own words, or send a photo, or ask about your particular plant and what you have written about it.
- What if what I am seeing is not on the list?
There is a line under the list that says None of these? — it takes you to the chat, where you can describe it in your own words or send a photo. The list only holds what somebody has written up for that crop, and gardens produce plenty that nobody has.
- Why does it name the university that wrote it?
Because you should be able to check us. Every cause we show is somebody else's work — a university extension service, an agriculture department — and we show their words rather than our summary of them, with their name and a link. If you want to know how far to trust a sentence, the fastest way is to read who wrote it.
- Does it work without a signal?
Not yet. This one needs a connection, because the write-ups live on our side rather than on your phone. If you are out at the far end of the garden with no bars, take a photo now and open the guide when you are back in range — the photo is saved on your phone either way.
- The weather days used to show a percentage. Now some of them show inches.
Because a percentage was never the number you needed — changed 31 August 2026.
A 90% chance of a passing shower can wet nothing. A 40% chance of a thunderstorm can drop half an inch. If you are deciding whether to water, the chance tells you almost nothing and the amount tells you everything.
The forecast has carried the amount all along and no screen read it. We were showing you the chance because that is what got wired up first.
Now: if the forecast gives us an amount, you see the amount — 0.4 in, or a trace. If it does not, you still see the chance, because that is all anyone has for that day.
A blank means the forecast said nothing, not that no rain is coming. We never turn we were not told into a zero.
Shipped 2026-08-25. Last checked 2026-09-01. Something here wrong or out of date? Tell us — support docs going stale is a bug.