To work out notions of global causality properties of minimizers in the CFS framework seems to be one of the most interesting things to study on the abstract side of things. The intuition that closed timelike curves (aka timetravel) should be impossible in the CFS framework stems from the fact that by cutting all closed timelike curves in a Lorentzian manifold, the number of timelike separated points drops significantly. Given that the Causal action principle aims to minimize the amount of timelike separation in the manifold one would expect that spacetimes emerging in the continuum limit from CFS should not contain closed timelike curves. Here you can find a short proposal I wrote for a proposal writing workshop.