[1]
“Detection and Mitigation of Prompt Injection and Memory Poisoning Attacks in Agentic AI Systems”, AMAR, vol. 4, no. 2, pp. 653–668, Feb. 2026, doi: 10.66021/.