I divide projects in "care code" and "care result". I am buiding random tools all the time for which I dont check the code. For my long term project I do check the code.
This is the best I came with so far, but I have to admit how slow my main projects move, if at all.
Yeah I’ve tended to align with this split as well. I will say that there has been a huge productivity boost from making these random throwaway tools composable because then I can chain them together via old-fashioned shell wizardry or having agents use them. Also have found it more repeatable and reliable to ask an agent to build me a tool to do a task (above some complexity threshold) vs asking it to do the task directly.