This will sound extremely stupid, but I have a skill that I use after letting an agent write code for a new feature. This skill creates a copy of the base branch and shows me each of the diffs that matter (outside of imports and other non-important boilerplate) and then I type it on an editor.
I don't use it too often because of the time it takes, but for critical work it is really good.
I have tried to come up with other equivalent ideas, like asking questions about a specific aspect of a feature, or showing the diff with some gaps to be filled, and even though they help I am still on the look out of methods to "imprint" the understanding of features into our brain.
Does anyone know any good papers or studies that could be useful in this research?
That’s so interesting. I remember trying to copy proofs in school out by hand but it did not work for me because - I think - I didn’t engage with the material as much as I should’ve. How do you ensure that you’re not simply blindly typing the code out and actually understand what’s going on?
Well, I only do it in order to try to understand it but there is no way to make sure you are understanding. But most of the time, if I am typing it's almost automatic that I pay attention to it. Not sure if it would work for others tho.
When you are writing code manually, actually you are thinking about what you are dong, what you will build actually, what are your or users' real requirements, etc.