Path-Aware Defense Against Prompt Injection Attacks in Retrieval-Augmented Large Language Models. (2026). Journal of Advanced Artificial Intelligence Research, 1(1). https://www.jaair.org/index.php/home/article/view/132