Phase 4 · Prove · Point 9 of 12
Security architecture for agents
Least privilege, secrets management, egress control and sanitization against injection and exfiltration.
Read the full lesson and take the quizTraditional application security assumes code executes what developers wrote. Agentic systems add a reasoning layer that inputs can manipulate: it can be tricked into revealing data, misusing tools, or taking actions its user should not be able to take. Security has to cover both the model and the tools it can call.
Security needs to be designed as a first-class component of the agent platform, not a patch.