[guided]The point of this step is that total boundedness survives passage to closure. We start from the precompactness of $A$ and translate it into total boundedness using [citetheorem:8517]. Thus, for every radius $\delta > 0$, finitely many $\delta$-balls centered at points of $A$ cover $A$.
Let $C := \overline{A}$, with restricted metric $d_C(x,y) := d(x,y)$ for $x,y \in C$. To show that $C$ is totally bounded, fix $\varepsilon > 0$. We deliberately cover $A$ by balls of radius $\varepsilon/2$, not $\varepsilon$, because a point of $C$ may not lie in $A$; we first approximate it by a point of $A$, and then use the finite cover of $A$.
Since $A$ is totally bounded, choose $a_1,\dots,a_N \in A$ such that $A \subset \bigcup_{i=1}^{N} B_X(a_i,\varepsilon/2)$. Now take any $x \in C$. By the definition of closure, every open ball around $x$ intersects $A$. In particular, there exists $y \in A$ such that $d(x,y) < \varepsilon/2$. Since $y \in A$, the finite cover above gives an index $i \in \{1,\dots,N\}$ such that $d(y,a_i) < \varepsilon/2$. Applying the triangle inequality gives $d(x,a_i) \leq d(x,y) + d(y,a_i) < \varepsilon$. Thus every $x \in C$ lies in one of the finitely many balls $B_X(a_i,\varepsilon)$. Since $\varepsilon > 0$ was arbitrary, $C=\overline{A}$ is totally bounded.[/guided]