Explore prompt injection defenses for educational LLM tutors, balancing security, usability, and latency. Understand trade-offs and practical guardrail strategies for robust AI learning environments.