It would have to explain all the coupling parameters and all the mixing angles of the standard model.
It would have to explain why there are exactly 3 families of fermions, (electron, tau, mu) and corresponding quarks (u,d; c,s; t,b) --families of particles.
It would have to explain not-so-well understood components of gravity (vacuum energy and dark matter)
There is a swathe of accidental events that would not necessarily have to be included.
Example: Why did a Mars-sized planetoid collide with Earth circa 3.9* billion years ago?
Why did the Permian, Cretaceous, etc, extintions take place? Why am I here, drinking some wine, talking to you? (that's certainly part of everything),
etc.
Those are considered historical contingencies. Other cosmological problems you would be forgiven for not being able to explain, like matter-antimatter asymmetry, details on CMB etc.
* 4.5? I don't remember.
I almost forgot: It wouldn't hurt to know (if possible) why all dimensions are unobservable except for 4.
All in all, that's a decent summary of what scientists understand by "theory of everything".