[8note]

I don't see this listed yet, so do the same thing you'd do at work: keep a prioritized backlog, and do project planning to decide what to work on next.

When you see an architectural defect, write it into your backlog, then flesh it out as part of grooming on your backlog for whether it's worth doing or not.

You might want to take a course/training on product or project management so you can practice switching hats