When we have a rigid body, the rigidity constraint allows us to write the trajectory $\mathbf{r}_i$ of the $i$-th particle as
$$\mathbf{r}_i(t) = R(t)\mathbf{b}_i + \mathbf{w}(t),$$
where we are using a $\mathbf{b}_i$ as the initial position of the $i$-th particle with respect to one stationary frame and where $R(t)\in SO(3)$ is a path of rotations.
In that case we have the velocity
$$\mathbf{v}_i(t)=R'(t)\mathbf{b}_i+\mathbf{w}'(t)=R'(t)R^{-1}(t)(\mathbf{r}_i(t)-\mathbf{w}(t))+\mathbf{w}'(t),$$
thus recognizing that $R'(t)R^{-1}(t)\in \mathfrak{so}(3)$ and remembering this is the Lie Algebra of antisymmetric $3\times 3$ matrices, there are numbers $\omega_1,\omega_2,\omega_3$ so that
$$R'(t)R^{-1}(t)=\begin{pmatrix}0 & -\omega_3 & \omega_2 \\ \omega_3 & 0 &-\omega_1 \\ -\omega_2 & \omega_1 &0\end{pmatrix},$$
we are able to write the action $R'(t)R^{-1}(t)$ as $\omega(t)\times$ where $\omega(t)=(\omega_1(t),\omega_2(t),\omega_3(t))$. Thus we find
$$\mathbf{v}_i(t)=\omega(t)\times \mathbf{r}_i(t)-\omega(t)\times \mathbf{w}(t)+\mathbf{w}'(t),$$
and we recognize $\omega(t)$ as the angular velocity. This constructions is quite nice as to explain where the angular velocity comes from and its properties.
On the other hand, I don't know how to properly identify what reference frame the components $\omega_i$ are with respect to. The obvious thing would be to say that it is with respect to the stationary system, but I don't know how to justify this.
Are these components really with respect to the stationary frame? If they are, how do we justify it? If not, these components are with respect to what frame?