It has been designed to be easily modifiable, so you can create your own tree by changing the deffact called 'Nodos'.
By default, it shows the most likely node of the desired* levels of the tree, but you can show the rest of the nodes probability (of those levels) by commenting specific lines of the code (which have been marked).
(The desired* levels to show are contained in the fact 'toString2' of the rule 'inicio'.)
.
With 'ClipsSalud' you can solve any compensated tree like this one (given as default):
PS: It's not designed for unbalanced trees, so it may not work. Also, it doesn't use classes.
(It's commented in Spanish!)