Module 6, Agent Governance and SecurityOutline
Prompt injection and untrusted instructions
Everything an agent reads is data. Instructions planted in a README, an issue or a dependency, and the defenses that hold.
This lesson is part of the paid course
The course is not on sale yet. If you already have a license key, unlock it here.
Questions on this lesson
Questions and answers are open to learners who unlocked the course. Unlock the course to read them and ask your own.