Stokes' theorem is valid everywhere, in the sense that if you have a field $\vec{F}$ defined everywhere, then $\int_{\partial S} \vec{F}\cdot d\vec{s}=\int_S (\vec{\nabla}\times\vec{F})\cdot d\vec{a}$. The problem is, what is your field $\vec{F}$ that you want to define? Since the value of the force $\vec{F}$ doesn't depend on just your position in space but on the particle's velocity as well, it's not possible to unambiguously define a force field at every point in space, since once you get to that point, your particle might have any velocity (and thus any possible force).
I'll add that in the case of the magnetic field itself, you do have a field $\vec{B}$ defined everywhere in space, and so Stokes' theorem applies perfectly well to $\vec{B}$: $\int_{\partial S} \vec{B}\cdot d\vec{s}=\int_S (\vec{\nabla}\times\vec{B})\cdot d\vec{a}$. Combining this with the fact that $\vec{\nabla}\times\vec{B}=\vec{J}$, you get Ampere's Law. So it's not like Stokes' theorem fails for magnetism, it's really just that there is no force field $\vec{F}$ to apply Stokes' theorem to.