The correct answer is Manual peer inspection of source code by developers.

Manual peer inspection of source code by developers is a human driven process where developers read and discuss each others code to find logic errors, security issues, maintainability problems, and opportunities for improvement. It emphasizes conversation, shared understanding, and judgement rather than automated checks. When people refer to code review in a development context they are normally referring to this practice.

Automated static code analysis tools are automated programs that scan source code for known patterns and common defects. They are useful complements to a review but they do not replace the human judgement and collaborative discussion that a code review implies.

Compilers run with extra diagnostic flags can surface syntax and some semantic issues during the build process. They perform automated checks at compile time but they do not involve peer inspection of the source and so they are not what is typically meant by code review.

Runtime dynamic analysis and fuzz testing tools exercise the running program to find runtime errors, memory issues, and unexpected behaviors. These dynamic testing techniques focus on execution and inputs rather than on human inspection of the source code and therefore they are not the same as a code review.

Full AWS Practitioner Certification Question