Jon Udel describes Bram ("Bram runs agents mindfully"), a "desktop app with three panes: a terminal where you use Claude Code and/or Codex, an agent pane that embodies a workflow (rendered by XMLUI), and an app pane that hot-reloads the app you are developing." This is basically also how I've set up my mixture of a VPS, VS Code and Claude in my own development workflow, so I can see the appeal. Why is this such a good approach? "It encourages agents to enact a git/gh-centric workflow that makes otherwise chaotic agent-assisted development feel safe, orderly, and accountable," says Udell. What that means is you can manage code sharing and collaborative development while allowing your agent to take care of the complexities of doing that. It also "helps you think clearly about the work you are doing, and proceed in well-defined chunks and sequences." Developing this way hasn't redefined my undertsanding of thinking and learning, but it has helped me understand them to deeper degrees of depth. Yes, it's like playing with fire.
Today: Total: [Direct link] [Share]
