16767

Low-metallicity massive stars may have reionized the universe at z > 6. Yet comma their contribution to the ionizing photon budget remains unconstrained and can only be estimated from stellar models. The predicted ionizing flux is highly sensitive to theoretical mass-loss rates and metal opacities comma but these have never been observationally determined below the metallicity of the SMC (20% solar). Observations of individually resolved comma metal-poor O stars spanning a range of masses and evolutionary stages are needed to determine the contribution of high-redshift galaxies to cosmic reionization. Sextans A is the only nearby dwarf galaxy substantially more metal-poor than the SMC that hosts more than a few known O stars. Five have already been observed by COS and two more will be observed by ULLYSES comma but these do not sample the full range of spectral types and luminosity classes. We propose new G130M and G160M COS observations of two late-O (super)giants that sample new parameter space. For the first time at 6% solar metallicity comma the combined new and archival COS data will enable us to (1) empirically determine the photospheric metal opacities that set ionizing photon production (2) measure mass-loss rates as a function of spectral type to improve the purely theoretical mass-loss prescriptions that underlie stellar evolution models and (3) produce new stellar spectral templates tailored to the FUV observations that will be ideally suited to modeling the ionizing spectra of galaxies during the epoch of reionization. The proposed observations are urgently needed to interpret upcoming JWST campaigns at high redshift and can only be done with HST-COS.