Understand this story
In the article "How the Claude Code team does code review," Addy Osmani, Morgan Westlee Lunt, and Qing Wang propose that code review should not only ask "did a human read the code," but should focus on what could go wrong with this change and which kind of check should catch it. The article states that code review thus shifts from a gate at the end of a PR to an ongoing process, with most changes already reviewed and fixed at least once by Claude before a human looks at them. This view is currently reported by Simon Willison on October 17, 2026, and there are no subsequent developments yet.
AI-generated from public reports
Latest development · Report
How the Claude Code team does code review: let Claude review first, then hand it to humansDevelopments
1 confirmed developments
Reports and discussion timeline
Browse reports and discussion by time. Their evidence levels remain distinct.
Oct 17
Simon Willison
How the Claude Code team does code review: let Claude review first, then hand it to humans
In "How the Claude Code team does code review," Addy Osmani, Morgan Westlee Lunt, and Qing Wang argue that instead of asking "did a human read the code," we should ask "what could go wrong with this change, and which kind of check should catch it." Code review thus shifts from a gate at the end of a PR to something continuous, with most changes already reviewed and fixed at least once by Claude before a human looks at them.