Enrico,

here are the steps (check for typos as I was typing all the below from
memory):

> 1) read in a mmcif file (which I think I can do following the tutorial);

pdb_inp = iotbx.pdb.input(file_name = "4znn.pdb") # can be PDB or mmCIF
model = mmtbx.model.manager(model_input = pdb_inp)

> 2) trim the solvent and hydrogen atoms (probably I can do it following
> the tutorial);

model_no_H = model.remove_hydrogens()

model_no_water = model.remove_solvent()

though better:

selection = model.selection("not (element H or element D or water)")

model_trimmed = model.select(selection)

> 3) run reduce2 to get hydrogens in the nuclear positions (no clue);

Run
mmtbx.reduce OR
mmtbx.reduce2 OR
mmtbx.hydrogenate

to add H. I have no idea about the difference between the three commands
(perhaps just the mess that someone has to finally get brave enough to
clean!), but all three are supposed to add H.

Also, you should be able to access the above calls programmatically, at
Python script level, not running in the CL.

> 4) get the full content of the unit cell (no clue).

Do you mean expand to P1 (if original symmetry is not P1)? If so, then

hierarchy = model.get_hierarchy().expand_to_p1()

Happy to answer questions!
Pavel