When crow connects to norns, we ‘reset’ the crow virtual machine, stopping any currently running script. This is by design, and was implemented a while back because it became very confusing to people when they tried to run a norns script with crow integration and it was partially still doing the ‘crow-script’ thing, plus whatever the norns script added on top.
We don’t clear the script on crow when connecting to norns, simply deactivating it instead. That script is still available though, hence my suggestion of simply running crow.send "init()" which tells crow to run it’s internal init() function that the crow script should define.
One thing to note, whenever you load a new norns script, the crow environment is reset. This means you could have a script that just executes the above line, but as soon as you reload another script to have norns do anything, it will disable crow again.
From your message it sounds like the above is not working correctly, but I’m having a hard time understand what is happening vs what you’re expecting. If you could run through what happens on crow before you attach it norns, then when you connect to norns, and again when you run the crow.send "init()" command. If you could post the crow script you’re using that’d be super helpful too. I looked at Cyrene and nothing in there suggests it would break this functionality.
I know you say it’s not a big issue for you, but I think it should work, and perhaps there is a tiny fix that needs to be made to make it act as expected.