Identifying Security Boundaries

A security boundary separates code, data, or users with different levels of trust. Every boundary should have an explicit verification step before data crosses it.

Start with trust

Treat input from browsers, third-party services, and other internal services as untrusted until the receiving system validates it for the operation being performed.

Untrusted request -> validation -> authorization -> protected operation