AI coding assistants have transformed software development by speeding up code writing, but they haven’t solved the challenge of code review. Tools like Greptile and Cursor’s BugBot can catch bugs but fail to capture the collaborative aspect of reviews. Effective PR reviews are about communication—ensuring the team understands the changes and their implications. The author, a CTO, emphasizes the need for a workflow that allows for quick yet thorough reviews. The proposed setup requires a compatible IDE, the GitHub CLI, and optional tools for better diff visibility. The author suggests using AI to generate a structured review plan rather than the review itself. The workflow has several steps: identifying relevant PRs, examining the changes, assessing impacts, and preparing feedback. Each step includes specific commands and considerations, like checking for active dependencies and noting potential issues. The AI also helps compile suggested comments in a user-friendly tone, streamlining the review process. Once the review is drafted, the author can refine it before submission, ensuring that the final output reflects their insights. This iterative process not only saves time but also enhances future reviews. By automating the more tedious parts of the review, the author maintains control over the final comments, ensuring that human judgment is central to the process.