Yes, Rama probably isn’t semantically comparable to one single monad.

I was talking about the do notation as a way to sugar the syntax of cps monadic operations into a flat, imperative syntax. This is exactly what Rama is doing.

If you look at a tutorial of what haskell do-notations desugar into, you’ll find the same cps stuff described in this article.