I have a github action that runs hourly. It pulls new issues from sentry, grabs as much json as it can from the API, and pipes it into claude. Claude is instructed to either make a PR, an issue, or add more logging data if it’s insufficient to diagnose.

I would say 30% of the PRs i can merge, the remainder the LLM has applied a bandaid fix without digging deep enough into the root cause.

Also the volume of sentry alerts is high, and the issues being fixed are often unimportant, so it tends to create a lot of “busy work”.

EDIT: It did let me in, but I don't know why it took so long.

I've worked on teams where there's been one person on rotation every sprint to catch and field issues like these, so taking that job and giving it to an AI agent seems like a reasonable approach.

I think I'd be most concerned about having a separate development process outside of the main issue queue, where agents aren't necessarily integrated into the main workstream.