Anthropic just gave its command-line tool Claude Code a serious upgrade. Meet Agent View — a built-in task panel that lets developers schedule, suspend, and manage multiple AI agents all from the same terminal window.
Before, running several agents at once meant juggling a bunch of terminal tabs or tmux windows. Now you just type claude --bg [task] to dispatch a new agent to work in the background, or use the /bg command to push your current session into the background.
Hit the left arrow key or type claude agents, and the terminal switches to a global list. Each row shows the task's current status, its last reply, and how long it's been idle. If an agent is stuck waiting for instructions, you can reply right there in the panel without diving into the full session.
This concurrency model is a huge productivity boost for solo developers. You can fire off a bunch of refactors or bug fixes at once, then come back to the list to review all the pull requests they've generated — no more staring at a single task until it's done. For now, the feature is available as a research preview to Pro and above subscribers and API users.