https://en.wikipedia.org/wiki/Theory_of_everything

In order for a theory to qualify, it has to have been extensivly tested and have a strong consensus in its favor. It also must explain everything fundamental that is in principle explainable. (It doesn't have to answer philosophical questions like "why is there something rather than nothing", since that may not be answerable.) In other words, there shouldn't be any physics question where we would not know how to compute an answer. (Even if the computation itself would take a very long time.)

There will of course be many emergent phenomena that we still don't have good models for. (e.g. we don't have a good way to predict many chaotic systems, even if we understand the underlying particle physics.)