Comparative Empirical Evaluation of Prompting Strategies for Large-Language-Model Web-Navigation Agents on the WebArena Benchmark. J. Sci. Innov. Soc. Impact. 2026;2(4):25-36. doi:10.71222/jbq63281