Four Things to Note About a Recipe After Cooking It
Recipe apps record what the publisher wrote. The useful information is what happened when you made it, and there is usually nowhere to put it.

The short version
Record four things after cooking something: how long it really took, what you changed, who liked it, and whether it works made ahead. All four are about your kitchen rather than the recipe, which is why no recipe site holds them, and they are what turns a saved collection into a usable one.
Every recipe app stores what the publisher wrote: ingredients, method, an optimistic time, a serving count. All of it is available on the original page, which means the app is mostly a copy.
The information that would actually make a collection useful does not exist anywhere yet, because it is generated by you, in your kitchen, the first time you make the thing.
Four fields, recorded once
How long it actually took. Published times assume a prepped kitchen, no interruptions, and a cook who has made it before. The honest number is frequently half again as long, and knowing your number rather than theirs is what makes a Tuesday plan survive contact with a Tuesday.
Record it the first time and never think about it again.
What you changed. Almost nobody makes a recipe exactly as written the second time. Half the garlic, no coriander, an extra ten minutes in the oven, swapped the cream for yoghurt. These adjustments are the difference between a recipe that works in your house and one that does not, and they are held entirely in memory, which loses them within about two months.
One line. “Half the chilli, cook 10 min longer.”
Who liked it. The highest value field and the one no software provides, because it is about the people you feed rather than about the food. A recipe that a specific person asked about again is not the same object as a recipe you enjoyed. When you are deciding what to cook for those people, that is the only field that matters.
Whether it works ahead. Discovered by accident, forgotten immediately, and decisive for whether something is useful on a weeknight or only on a Sunday. Can it be assembled in the morning? Does it reheat? Does it freeze?
Why this cannot live in the recipe
The instinct is to write these notes into a copy of the method. It does not work, for a practical reason: notes buried inside a wall of text are invisible at the moment you are choosing what to cook, which is when you need them.
They have to be fields, on the card, visible in the collection. The whole point is that scanning twenty cards should tell you which of them take under half an hour and which ones your household actually asked for again. Buried in paragraph six, they tell you nothing until you have already committed.
The compounding effect
This is a habit with an unusual payoff curve. The first month it feels like admin for no reward, because a collection of six annotated recipes is not much better than a collection of six recipes.
At about thirty, it changes. You have a group of things known to work, with real times, adjusted to your taste, tagged with who liked them. That group answers “what shall we have” faster than any recipe site can, because it is drawn from a set where everything is already known to succeed.
Which is what everyone thought they were building when they started saving recipes in the first place.