Oh, I should say, the cluster scripts functionality I added to crmsh was meant to eventually be able to implement the collection stuff in hb_report. Right now, the cluster health command calls a cluster script which in turn calls hb_report as part of its execution, but in theory it could do everything that hb_report does by itself. Cluster scripts can be written in any language, but right now the API that crmsh expects them to implement is somewhat complicated.. I'm hoping that I can simplify that so that writing new cluster scripts gets easier. Since it's the first iteration of the idea and I was figuring out just what the idea was while I was implementing it, the cluster scripts user interface is not perfect. But I have some ideas for improving it which would make it a lot more usable I think.