Let $b_1,\dots,b_n$ and $c_1,\dots,c_n$ denote the diagonal entries of $K_0$ and $K_1$ respectively. As I explain in my post here, there exists a permutation matrix $P$ such that
$$
PAP^T = \pmatrix{A_1 \\ & \ddots \\ && A_n},
$$
where
$$
A_k = \pmatrix{0 & 1\\-b_k & -c_k}, \quad k = 1,\dots,n.
$$
Thus, $A$ is Hurwitz if and only if each $2 \times 2$ $A_k$ is Hurwitz. We quickly conclude that every $A_k$ must be Hurwitz because the positivity of $b_k,c_k$ ensures that $A_k$ has negative trace and positive determinant.
Alternatively, because the block-entries of $A$ commute, we have
$$
\det(A - \lambda I) = \det\pmatrix{-\lambda I & I\\ -K_0 & -K_1 - \lambda I} =
\det[(- \lambda I)(-K_1 - \lambda I) - (I)(-K_0)]\\
= \det[\lambda^2 - K_1 \lambda + K_0]\\
= \det \pmatrix{\lambda^2 + c_1 \lambda + b_1\\
& \ddots \\
&& \lambda^2 + c_n \lambda + b_n}.
$$