With the rise of AI coding, how do you currently handle issues that show up in stack traces? Do you write issues manually, input them into Sentry, or do you prefer to fix the bug directly without creating a ticket?
2 Answers
I usually just let Sentry do its thing. It automatically creates issues with the relevant stack traces included. It saves a lot of hassle since you're getting the context straight from the environment without having to do any extra work.
My workflow is usually Sentry → triage → GitHub issue if it looks serious. I don't create a ticket for every stack trace. I first check how often it happens, how many users are affected, and if it’s already on the radar. If it’s something actionable, then I create a detailed issue instead of just dumping the trace into GitHub. Plus, I’ve been asking AI to analyze Sentry logs to see if there’s something genuinely wrong before moving forward.
Totally agree! I sometimes confuse it with old practices and think about pasting things into Sentry, but it’s really not necessary.