My train of reasoning (well, not even that, more like "translating") so far:
Let $C$ be some constant. $\|\vec {v(t)}\|=C$, so $\frac {d\|\vec {v(t)}\|} {dt}=0$.
But, where to go from here? I don't know how to express $\frac {d\|\vec {v(t)}\|} {dt}=0$ in terms of $\vec a$, aside from just flatly saying "$\mathrm{comp}_va=0$", which I feel doesn't sound very smooth or explanatory.
With some help from Semiclassical, I figured it out:
Let $C$ be some constant; we know $||\vec {v(t)}||=C$ and $||\vec{v(t)}||^2=\vec v • \vec v$ so $\vec v • \vec v=C^2$ and $$\frac {d(\vec v • \vec v)} {dt}=2(\vec v • \frac {d \vec v} {dt})=0$$Therefore, $$\vec v • \vec a=0$$