Time-dependent density functional theory (TDDFT) simulations of transient core-level spectroscopies require a balanced treatment of both valence- and core-electron excitations. To this end, tuned range-separated hybrid exchange–correlation functionals within the generalized Kohn–Sham scheme offer a computationally efficient means of simultaneously improving the accuracy of valence and core excitation energies in TDDFT by mitigating delocalization errors across multiple length-scales. In this work range-separated hybrid functionals are employed in conjunction with the velocity-gauge formulation of real-time TDDFT to simulate static as well as transient soft x-ray near-edge absorption spectra in a prototypical solid-state system, monolayer hexagonal boron nitride, where excitonic effects are important. In the static case, computed soft x-ray absorption edge energies and line shapes are seen to be in good agreement with experiment. Following laser excitation by a pump pulse, soft x-ray probe spectra are shown to exhibit characteristic features of population induced bleaching and transient energy shifts of exciton peaks. The methods outlined in this work therefore illustrate a practical means for simulating attosecond time-resolved core-level spectra in solids within a TDDFT framework.