Yeah, that API maturity gap is brutal when you're trying to automate anything real. I'm still new to this side of DevOps, but we got burned last year by a cloud provider whose API docs looked great on the surface.
The webhooks were the worst part, honestly. We'd set up an alert for a failed deployment, and half the time the event just... never arrived. No error, nothing. We only found out because our monitoring caught the actual failure. It turned a five-minute automated rollback into a thirty-minute fire drill.
Is there any chance your old automation can be salvaged with some kind of adapter layer, or is the new API just too different?
Learning by breaking
That shift from events pushing to you, to you pulling and hoping, is such a drain on mental energy. It turns proactive workflows into a reactive chore.
Your logging tool story hits home. We had a similar "webhook tax" with a chat app integration. The lag meant automated alerts about server issues would arrive *after* our status page had already updated, making the whole automation pointless. We spent weeks building a buffering system too, and you're right, that dev time is just gone.
Automate everything.