It can be very useful to let Gauss write out MCParticles in the Generation directly (without the detector simulation), which then can be analysed. This can be achieved the following way (write this at the end of your version of Gauss-Job.py):

gtos = GenerationToSimulation()
gtos.SkipGeant = True
gs = GaudiSequencer("GenMonitor")
MCTruthStream = OutputStream("MCTruthStream")
MCTruthStream.ItemList += [ "/Event/Gen#1", "/Event/Gen/HepMCEvents#1", "/Event/MC#1", "/Event/MC/Header#1", "/Event/MC/Particles#1", "/Event/MC/Vertices#1" ]
MCTruthStream.Output = "DATAFILE='MCTruth.sim' TYP='POOL_ROOTTREE' OPT='RECREATE'"
gs.Members += [ SimInit(), gtos, MCTruthStream ]

The root file can then be run over with an MCTupleTool in DaVinci (with certain restrictions, i.e. only tools which only access 4-vectors will run…)

You can run EvtGen standalone, letting the particles decay at rest. This is very fast, as no underlying event has to be generated. A prototype file could look like (Gauss v41r1):

#Event Type: 11114001
from Configurables import Generation
Generation().EventType = 11114001
from Configurables import StandAloneDecayTool
Generation().SampleGenerationTool = "StandAloneDecayTool";
Generation().addTool( StandAloneDecayTool )
Generation().StandAloneDecayTool.ProductionTool = "PythiaProduction";
 
from Configurables import ToolSvc
from Configurables import EvtGenDecay
 
ToolSvc().addTool( EvtGenDecay )
ToolSvc().EvtGenDecay.UserDecayFile ="$DECFILESROOT/dkfiles/Bd_Kstmumu=DecProdCut.dec"
Generation().StandAloneDecayTool.SignalPIDList = [ 511,-511 ]

These both things (running EvtGen standalone and writing out MCParticles) can of course be combined.