本文旨在探討利用AI Agents應用程式開發，並聚焦在簡化MASEM矩陣的原始資料建構流程。透過設計符應建構矩陣的指令與簡易的設定，且利用一篇MASEM論文進行實測，經實測結果發現，利用AI Agents可協助從事MASEM研究者快速且正確地擷取效果值，展示了AI Agents在MASEM矩陣建構的效益。本文認為在AI技術持續進化下，研究者應保持更開放的心態，透過人機協作，將繁雜的事交給AI，將精力集中於研究結果的洞見中，在研究者持續開發下，學術研究工具將邁向更智慧化的時代。

This study aims to explore the development of AI agent applications, with a specific focus on streamlining the raw data construction process for MASEM matrices. Through the design of instructions tailored for matrix construction and simplified configurations, coupled with empirical testing using a published MASEM study, the results demonstrate that AI agents can assist MASEM researchers in rapidly and accurately extracting effect sizes. This research showcases the efficacy of AI agents in MASEM matrix construction. The findings suggest that as AI technology continues to evolve, researchers should maintain a more open mindset and embrace human-AI collaboration, delegating complex tasks to AI while concentrating their efforts on deriving insights from research outcomes. With continued development by researchers, academic research tools are advancing toward an increasingly intelligent era.