The morning starts with going through the logs — the errors that piled up overnight. Nothing complicated. You read the errors, you go into the code, you see why a wrong parameter came in here, and over there why a parameter is missing instead. You commit. Note to self: check again in a couple of hours.
At the meeting you describe what you spent your time on — and the response is an idea. Maybe you need a skill? Look: Claude Code will do all of this for you, and you just review the merge requests. We schedule the skill with cron and that’s it.
Cool, right? Cool.
And it really works
Let’s roll it out to the neighboring repositories.
Wait, hold on — here’s an error I’ve seen before. Oh, I get it: the skill didn’t understand that this isn’t a bug, it’s a new feature, and it’s suggesting we delete it.
Sec — there was a test right here. Where did it go? Who merged the test deletion? It was actually being used — it was just marked as unused.
And over here the query got “optimized.” But it was written that way on purpose — the client’s database isn’t on the new version yet.
We should track this
We need to add all of this to the skill. The thing about the not-a-bug. The thing about the load-bearing test. The thing about the deliberately un-optimized query.
Let’s create a place where we’ll track all these issues.
A backlog, of sorts. For the problems with the skill.
We’ll call it: Skill Issues.