It's been running in production since then, tracking JS, PHP, C#, and Java errors. Nothing fancy - error grouping, stack traces, trends, uptime checks, alerts. The kind of stuff you actually look at.
Now there are dozens of monitoring tools, but most of them have grown into expensive, complex platforms. Muscula stayed small on purpose. It just does the job.
What's new: I've recently added a CLI and an MCP server. The CLI lets you browse errors from the terminal (`muscula errors "MyProject"`) and pipe output directly into Claude Code or any other tool: muscula errors "MyProject" | claude "anything critical here?"
What MCP integration does probably you know, you can just ask AI "fix all the problems" and it's fixing.
This is for people who want to know what's broken in their app without paying too much per service or wading through a complex UI.
I'm using it all the time, hopefully you'll find that useful too.
Stack: .NET MongoDB, React, Mobx.