In the light of recent advances in understanding the Auger process, the local electronic structural origins of selected core-valence-valence ($\mathrm{XVV}$) Auger line shapes are analyzed for a transition-metal silicide prototype, ${\mathrm{V}}_{3}$Si. We report for clean $p(1\ifmmode\times\else\texttimes\fi{}1)$-${\mathrm{V}}_{3}$Si(100) Auger spectra that include the region of the vanadium ${M}_{2,3}\mathrm{VV}$ and ${M}_{1}\mathrm{VV}$ and the $\mathrm{Si}{L}_{2,3}\mathrm{VV}$ transitions. We compare the measured line shapes to spectra we generated based on the muffin-tin local density of states (DOS) calculated self-consistently by Klein et al. Good agreement in both the $\mathrm{Si}pp({L}_{2,3}{M}_{2,3}{M}_{2,3})$ and the $\mathrm{V}dd({M}_{2,3}{M}_{4,5}{M}_{4,5})$ peak positions between experiment and calculation verified that the final-state hole-hole repulsion for $\mathrm{Si} ({U}_{\mathrm{pp}})$ and $\mathrm{V} ({U}_{\mathrm{dd}})$ are both \ensuremath{\sim}0 eV. Also, the $\mathrm{Si}{L}_{2,3}\mathrm{VV}$ spectrum resembles that of elemental Si in that the line shape is predominantly a self-fold of the $\mathrm{Si}3p$ DOS. However, an unexpected result is that the V spectral region above the ${M}_{2,3}\mathrm{VV}$ threshold possesses a broad (\ensuremath{\sim}30-eV-wide) intense feature that is not amenable to conventional interpretation in terms of the ${M}_{1}\mathrm{VV}$ transition or ${M}_{2,3}\mathrm{VV}$ double-ionization or plasmon-gain satellites. We attribute this observation to the presence of Fano autoionization emission associated with deexcitation of the resonant $3p\ensuremath{\rightarrow}3d$ transition. Supporting evidence comes from a comparison of our x-ray- and electron-stimulated Auger spectra, and to the line shape of the $3p$ loss spectrum. In addition, oxygen-dosing Auger and x-ray photoelectron spectroscopy experiments (0-20 L) (1 langmuir = 1 L = ${10}^{\ensuremath{-}6}$ Torr sec) indicate dramatic $\mathrm{Si}{L}_{2,3}\mathrm{VV}$ line-shape changes associated with oxidation, similar to that observed previously for ${\mathrm{Pd}}_{4}$Si. The initial oxidation rate is \ensuremath{\sim}${10}^{2}$ faster than that for elemental Si. We hypothesize that the dissociation of ${\mathrm{O}}_{2}$ is a rate-determining step in the oxidation of elemental Si, but is rapid at transition-metal sites in the silicides. Atomic oxygen then rapidly spills over to the neighboring silicon sites where oxidation subsequently occurs.