Figure out an end-to-end testing approach

[hbagdi]

decK v0.5.0 had a very simple bug that rendered the release unusable. This could have been easily could with an automated end-to-end testing setup, which we don't have at the moment. As more changes are introduced, manual testing is becoming more expesive and proving not to be sufficient. Figure out a path forward for how do we want to test decK in an automated fashion.