Issue with MadSpin decaying the Z in a BSM model

Dear MadGraph Experts,

We are using MadGraph+MadSpin to generate the following process:
pp > Z’ > t T’ with T’ > t Z (t and Z to anything)
Where Z’ is a Z-like heavy resonance and T’ a heavy vector-like quark.
The MG model used can be found here [1] with the param [2] and run [3] cards.
We use this proc card [4], the relevant line is (rho0 is the Z’ and topo is the T'):
generate p p > rho0 > topo tt
The MadSpin card can be found here [5], the ony modification with respect to the standard madspin card is:
decay topo > z tt / rho+ rho- rho0 , z > all all, ( t > w+ b, w+ > all all ), (t~ > w- b~, w- > all all)
When I run MG+MadSpin, MadSpin crashes with this error [6] and log [7]. The compilation and generation of full square matrix element (with decay) steps are very slow as well (~30 min for both of them).
See the shell output [6].

I tried removing the decay of the Z from MadSpin, like this:
decay topo > z tt / rho+ rho- rho0 , ( t > w+ b, w+ > all all ), (t~ > w- b~, w- > all all)
and things go smooth and fast (but of course I’d like to decay the Z with MS as well).
If I let only the Z decay, like this:
decay topo > z tt / rho+ rho- rho0 , z > all all
MS crashes again.

So it seems there is a problem in the treatment of the Z.
Is there a problem in my MadSpin setup?

[1] https://github.com/emanueleusai/Vp_to_VLQ/tree/master/onerho
[2] https://github.com/emanueleusai/Vp_to_VLQ/blob/master/Single/Zp_Zt_t_madspin/param_card.dat
[3] https://github.com/emanueleusai/Vp_to_VLQ/blob/master/run_card.dat
[4] https://github.com/emanueleusai/Vp_to_VLQ/blob/master/Single/Zp_Zt_t_madspin/proc_card.dat
[5] https://github.com/emanueleusai/Vp_to_VLQ/blob/master/Single/Zp_Zt_t_madspin/madspin_card.dat
[6] https://github.com/emanueleusai/Vp_to_VLQ/blob/master/out2
[7] https://github.com/emanueleusai/Vp_to_VLQ/blob/master/out

Cheers,
Emanuele