Anthropic's Claude model has played a pivotal role in aiding astrophysicists to craft the inaugural comprehensive ultraviolet all-sky map. Leveraging historical multi-wavelength astronomical observation data, Claude extrapolated roughly one-third of the sky area, addressing a 50-year void in ultraviolet surveys. This gap was primarily due to telescopes intentionally steering clear of the Milky Way's luminous zones. Remarkably, tasks that would have demanded weeks of meticulous manual labor were accomplished by Claude in a mere few days. It achieved this by coordinating multiple agents to manage the entire workflow, encompassing data gathering, calibration, cross-calibration, and extrapolation. The discrepancy between the extrapolated areas and actual measurements stands at about 10%, with each pixel meticulously annotated with its observed or extrapolated properties, along with the error margin. As AI agents progressively assume responsibility for such foundational scientific research tasks, human experts can redirect their focus towards guiding the overall direction and pinpointing critical issues.