For decades now, low-energy models of QCD have shown indications that a crystalline quark phase could be stable at high chemical potentials. Beyond models, however, there are numerous difficulties in investigating such a hypothesis in full QCD, such as the sign problem. Functional methods do not suffer from the sign problem, and thus, can access the high--$\mu$ side of the QCD phase diagram. The main tool used to look for signs of inhomogeneous/crystalline phases in low-energy models is the so-called ``stability analysis''. In this talk, I show how the standard stability analysis was generalised to be applicable in any theory, including QCD.