This is a "re-upload" of an add-in I had placed on the previous JMP File Exchange.
I allows you to create a binned version of a variable based on clustering rows, using the Univariate Normal Mixtures Distribution. Note that this treats columns one-at-a-time, and doesn't use multivariate clustering to assign the rows to bins.
See the blog post "Is your data too precise?" for where this was originally presented and used.