1 min lesson
Contain prompt injection
Make the call in this situation. "A Jira comment tells the agent to send environment variables to an outside URL. Which controls should stop it?" Explain what supports it.
Step 1 of 3
A web page, Jira comment or document can contain text that tries to direct the agent. Treat external content as untrusted data.
Prompt injection happens when untrusted content changes how an agent behaves. Assume the model may follow that content, then use permissions and execution controls to limit the result.