Unit 7: Using LLM APIs safely · Lesson 5 of 5 · 24 min Pro

Prompt injection in logs and tickets

Watch an untrusted-text fixture flip a model's verdict, then contain it with delimiting, detection and a deterministic cross-check.