When you carry out a SCF calculation, you diagonalize the Fock matrix, and you get the eigenvalues (orbital energies)
In the case of a CASSCF calculation, the orbitals of the active space are treated separately: in the subspace defined by these orbitals the Fock matrix is not diagonalized. The density matrix is the one that is diagonalized. That is why they don't have a definite energy and in the output you have zero energy.
Well, these orbitals don't have a well defined energy... So it is not possible, as far I know... The density matrix does not conmute with the Fock matrix.