I used helm charts https://github.com/converged-computing/flux-apps-helm. This mean I can use common base templates for the Flux MiniCluster (provided by the Flux Operator) and for each #HPC app I just need to expose the parameters in a template. See how simple it is for the user?