An ADE is where you run AI coding agents and steer the work. It’s slightly different to a traditional IDE, where you write the code and AI may help. In an ADE, agents do the work, and you direct and review it.
It lets you build across multiple Claude sessions, several branches, speak notes, create task cards and gives great visibility for how fast your parallel sessions are using your limit, stopping cleanly at your plan limit.
At the moment it’s for Claude Code and macOS 26 on Apple silicon only. It’s free to use and we’re keen to show it to people and get feedback from people currently building with Claude Code.