obra/superpowers@receiving-code-review

If code review feedback contradicts project patterns, investigate the suggestion against codebase realities first. Question the reviewer, formulate clear technical arguments explaining the conflict, and provide a reasoned response rather than blindly implementing changes that may break established architecture or performance goals.

The best way to pushback on questionable code review suggestions is to formulate clear, technical arguments grounded in codebase realities. Critically evaluate the feedback against project-specific constraints and technical correctness, then communicate the reasoned response rather than blindly accepting potentially incorrect or unnecessary changes.

AI agents should verify feedback before implementing code review changes to prevent blind acceptance of potentially incorrect or unnecessary suggestions. Rigorous technical evaluation against codebase realities ensures accuracy, maintains architectural integrity, and confirms relevance to project-specific constraints before any code is modified.

To systematically implement verified code review feedback, prioritize each item after completing technical evaluation against the codebase. Structure the implementation process to test each change individually, ensuring that every modification maintains technical correctness and adheres to project-specific constraints before moving to the next.