SpotAgent: Grounding Visual Geo-localization in Large Vision-Language Models through Agentic Reasoning | Read Paper on Bytez