code-review

Perform a code review for AI-specific risks by analyzing implemented features against the technical plan to identify AI risk patterns, validate architecture alignment, and provide actionable feedback for code quality and safety.

Audit code quality beyond basic linters by evaluating architecture alignment, business logic correctness, cross-cutting concerns, and security vulnerabilities to ensure comprehensive safety and technical plan compliance.

Verify implemented features match the technical plan by validating plan implementation, analyzing architecture alignment, and checking business logic correctness to catch deviations before merging.

Yes, the code review process assesses security vulnerabilities and performance by thoroughly auditing implemented features, checking cross-cutting concerns, and identifying AI-specific risk patterns.

Yes, you can use this to catch bugs after an AI agent implements a new feature by performing a deep review that checks for potential security vulnerabilities, logic errors, and deviations from the original technical plan.

Limitations exist when reviewing code for AI-specific risk patterns as the process focuses on architecture alignment, business logic, and technical plan validation, but does not replace runtime testing or basic linter execution.