Problem with effective operator for CLFV

Good afternoon everyone.

I have the following problem. I want to study the phenomenology of charged lepton flavor violation (CLFV) effectively using the "muon electron > electron electron" channel. To do this, searching a little here I found that using Lanhep I can implement contact interactions between fermions. I need to implement the model and get the analytical expression of the total Cross-Section through Calchep. The problem appears here. If I implement the following lagrandian in lanhep

lterm (gpLLLL/(4*Lambda**2))*(anti(e1)*gamma^mu*(1-g5)*e2)*(anti(e1)*gamma^mu*(1-g5)*e1)+AddHermConj.

the vertex not appear in the mdl files and the error "Internal error in vertex (E2, e1, E1, e1) (rs03)" is showing. But if I implement the following lagrangian

lterm (gpLLLL/(4*Lambda**2))*(anti(e1)*gamma^mu*e2-anti(e1)*gamma^mu*g5*e2)*(anti(e1)*gamma^mu*e1-anti(e1)*gamma^mu*g5*e1)+AddHermConj.

appear the vertex with the auxiliar fields in the mdl files, but with the same error "Internal error in vertex (E2, e1, E1, e1) (rs03)". Using this mdl files in Calchep and run the process e1 E2 -> e1 E1, the numeric and analytical expression for cross section is zero.

What am I not considering here? Thank you!