I decided to build this after an issue popped up with a recent Rails upgrade. We upgraded a Rails app to 8.1 and picked up a new default that changes the order after_commit callbacks run. Our upload code started telling the virus scanner about files before they were saved to S3, and the scanner just quietly dropped them. No exceptions were raised and uploads sat in "processing" for days. We should have had alarms in place, we know that now but I wondered if there could be some type monitoring that would just have noticed the change. I did some checking and the callback route went from roughly 27 requests an hour to zero when we deployed. That seems like a fairly easy change in behavior to automatically raise an alarm. So after a few weekends working with Claude I have a new tool.
A small agent sends one aggregated payload per minute. Then the server compares each release with the last one it cleared. It looks at things like error rates, latency per route, traffic that collapses, and missing job runs. The server then returns a verdict if the release cleared or not. Currently I have agents for Rails, Django, FastAPI, and ASP.NET Core.
It will work best for teams that deploy often and the app gets steady traffic. It's not meant to replace error monitoring tools like Sentry, but to work alongside them. Here's a recent release of its own: https://app.deployangel.com/share/KAaurrfRHcw43ZrjkWst2DLY?r...
The agents are MIT-licensed. It's completely free right now, with a free tier planned.
https://github.com/DeployAngel
I'm looking for a few teams to run it on a real app for a couple of weeks. I'd love to hear if it surfaces issues that other monitoring and alarms miss. Of course I'm hoping it proves to be useful, but if not, that's good information too.