Stage wants to fix a real problem: engineers merging code they don't actually understand. The new code review tool, which launched on Hacker News this week, positions itself as a human-first alternative to AI-heavy review workflows. Put people back in charge of understanding what's changing in their codebase.
But commenters spotted a gap. Stage surfaces what changed without adequately showing why or how. You need to see the ticket, the conversation, the decision trail. Without that context, you're just rubber-stamping diffs. Reviewers still need to jump to other tools for the full picture. That's not much of an improvement.
What the thread actually revealed is how engineers combine AI with reviews. One described reading the code first, then asking Cursor targeted questions about specific functions. Another treats AI as a "smarter grep" for their codebase. The pattern is clear. AI works best when it multiplies what you already understand.
There are no shortcuts to quality control.