Cosmic ray directions are somewhat anisotropic, but I think not enough that you'd be able to exploit this for shielding.
Here are the slides of a presentation which quotes an observed anisotropy of about $10^{-3}$. Here is a paper with some plots using data from IceCube and IceTop, which again shows anisotropies $\sim 10^{-3}$. Searching for 'cosmic ray anisotropy' will find a lot of results: I've just picked a couple here.
If you're thinking about shielding, then you also need to consider whether the object you are shielding ever might rotate with respect to an inertial frame, and deal with that it if will. Chances are if it's a spacecraft or attached to planet it will (but if it's attached to a planet you don't need to worry about the stuff coming up from underneath you so much).
Although these results apply to high-energy cosmic rays, the isotropy of lower energy cosmic rays tends to be higher. That's because their gyroradius goes down as their energy goes down, so wherever they started from their direction gets essentially randomised. From these lecture notes, a proton with energy $\approx 1\,\mathrm{TeV}$, and given a magnetic field of about $10^{-4}\,\mathrm{G}$ in the local interplanetary medium, the gyroradius is about $20\,\mathrm{au}$ (the radius of the orbit of Uranus is $\approx 20\,\mathrm{au}$). So only protons with energies significantly greater than that can maintain their direction once they enter the Solar system. From these lecture notes:
Observations of cosmic rays show that the arrival directions are relatively isotropic, and in fact the lower the energy (down to $10^{12}\,\mathrm{eV}$) the more isotropic the distribution of cosmic ray directions.