Attached is:
ex13-1a.inp, which is an example of the missing data report possibilities in Mplus
ex13-1b.inp a multi-group approach to assessing missingness differences across groups.
(he doesn't discuss missing by design issues, but this is pretty similar)
ex13-1c.inp additionally a covariance comparison as multigroup
ex13-2a.inp example of auxiliary variable syntax
ex13-2b.inp examplpe of using baseline as an auxiliry variable (note that you don't include baseline in the model)
ex13-2c.inp example using mplus shortcuts and new syntax
ex13-3a.inp Heckman model
ex13-3b.inp Diggle Kenward model
ex13-4a.inp Outcome dependent model
ex13-4b.inp parameter dependent model
ex13-4c.inp Known class approach to missingness
ex13-5.inp knownclass pattern mixture model
ex13-5b.inp knownclass pattern mixture alternate model
ex13-6a.inp symmetric factor pattern model
ex13-6b.inp symmetric factor pattern model for a (linear) growth model