Potential cause of harm to a system.

Injection causing unintended command execution.

Security model requiring continuous verification.

Resource owners control permissions.

Set of possible system attack points.

Inserting executable code into processing.

Boundary separating different trust levels.

Giving only necessary permissions.

Measure designed to reduce security risk.

Injecting scripts into web content.