If you’ve spent any real time with Claude Code, you know it’s insanely powerful—I use it for basically all my development now. But the more you lean on it, the more the gaps show up. The biggest one for me is reviewing AI-generated changes across a bunch of files. Claude can rip through edits fast, but when it’s time to actually verify what changed (and yes, you still need to do that), it gets painful—especially when I’m running multiple Claude Code sessions in parallel, which I do constantly. In this video I show a free tool I’ve been loving called Kintsugi. The key thing: it’s not trying to replace Claude Code (that would be a deal breaker for me). It’s an Agentic Development Environment (ADE) that sits on top as a visual layer, so I can see all my parallel agents in one place, track what’s still running vs. what needs review, and even comment on an agent’s plan of attack so it adjusts in real time. Then, instead of digging through logs and file-by-file diffs, I review changes in a much more PR-like experience. On top of that, Kintsugi can run different kinds of reviews—including security review—which makes a ton of sense coming from the Sonar team. It’s macOS-only and experimental, but it’s already making Claude Code workflows way cleaner.