This is from "Multivariable Calculus, Concepts and Contexts" by Stewart.
He says "The parametric equations for this curve are: $x=\cos t$, $y=\sin t$, $z=t$. This makes sense.
However, he then goes on to state that "Since $x^2+y^2=\cos^2t+\sin^2t=1$, the curve must lie on the circular cylinder $x^2+y^2=1$.
I understand the trigonometric identity, I just don't understand how we get to even consider $x^2+y^2=1$ since the vector function is defined as $r(t)=\cos t\hat{\bf i}+\sin t\hat{\bf j}+t\hat{\bf k}$.
Where does $x^2+y^2=1$ come from? How do we get to use that fact when $x=\cos t$ and $y=\sin t$?

It's a simple substitution. You know from trigonometry that $$\cos^2t+\sin^2t=1,$$ and you know from the parametric equation of the curve that $$x=\cos t, y=\sin t.$$ So for any point of the curve, it is true that $$x^2+y^2=1.$$ Just notice that you dropped the equality $$z=t,$$ meaning that you get more $(x,y,z)$ values that satisfy the equalities: you obtain a superset of the curve points. Indeed, the parametric equation corresponds to a linear entity (1D), while $x^2+y^2=1$ defines a surface (2D), which contains the curve.
COMPLEMENT
If you replace $z=t$ by $t=0$, which amounts to projecting the curve on the $xy$ plane, you get a circle centered at the origin, with radius $1$. Coming back to $z=t$, you will "unroll" the circle and extend it in the $z$ direction, while remaining on the cylinder, forming a spring (called an helix in geometry).