A common way to develop code (don’t blame the messenger) is to start the application watch for the logged errors, make changes and then restart the application again, looking for more logged errors.
That’s not how it’s done in the world of Clojure, not just because it’s bad, but because the process of starting any Clojure program is horribly slow. Instead, in Clojure normally you start the application and then connect with a terminal window running a Clojure REPL. For example, if you are running a Luminus server, when you start it, it will tell you what port the server is listening for REPL connections on.