Anthropic Gives Claude Code a Unified Dashboard to Manage All Your AI Agents

Anthropic has introduced agent view in Claude Code: one place to manage all your Claude Code sessions. Before this, running agents in parallel meant managing multiple terminal tabs, a tmux grid, and an overloaded mental ledger of what to tackle next.

With agent view, you can kick off new agents, send them to the background, and jump in only when Claude needs you. A single glance tells you which agents are waiting on input, which are still working, and which are done.

As someone who covers Claude Code regularly, this is one of those quality-of-life additions that addresses a real friction point. Running three or four parallel sessions used to mean context-switching across windows constantly. Agent view collapses that overhead directly into the CLI.

What Agent View Actually Does

Agent view works as a session roster inside the CLI. Each row shows a Claude Code session, its current state, its last activity, and whether it is working, waiting for input, completed, failed, idle, or stopped.

Agent view improves visualizing and interacting with your Claude Code sessions in the CLI. Press the left arrow from any session or run claude agents from the terminal to open it. Each row shows the session, whether it needs your input, the contents of its last response, and when you last interacted with it. Select a session to peek at the last turn.

Sessions keep running in the background without a terminal attached, which means closing the view doesn't kill your work. When you attach to a session, Claude posts a short recap of what happened while you were away. While attached, the session behaves like any other Claude Code session: every command, keyboard shortcut, and feature works.

Key Technical Highlights

Developers can open agent view from any session with the left arrow or start it directly with claude agents. They can also move an existing session into the background with /bg, launch a new background job with claude --bg "[task]", peek at the latest turn with the spacebar, reply inline, or attach to the full transcript with Enter or the right arrow.

A few other details worth noting from the official documentation:

• Agent view is a research preview and requires Claude Code v2.1.139 or later. Check your version with claude --version.
• The interface and keyboard shortcuts may change as the feature evolves, and administrators can disable agent view for an organization with the disableAgentView managed setting.
• Detaching never stops a background session: ←, Ctrl+C, Ctrl+D, Ctrl+Z, and /exit all leave it running. Use /stop from inside the session to end it.
• Press Enter or → on a selected row to attach, or press Alt+1 through Alt+9 to attach directly to the Nth session in the focused group.

Three Core Use Cases

Scaling concurrent sessions: Dispatch several ideas at once, each optionally paired with a skill, and return to a list of pull requests ready for review.

Managing long-running agents: PR babysitters, dashboard updaters, and other looping jobs show their next run time right in the list.

Navigating between separate sessions: When you're in the middle of a session, press the left arrow, start a related task or quick codebase question, then arrow right back into what you were doing. Peek shows the answer when it lands.

Use agent view when you have several independent tasks Claude can work on at once, such as fixing a bug, reviewing a pull request, or investigating a log. When you want to work through a problem together, attach to a session and use Claude Code interactively as usual. For workers that report results back into a single conversation, use subagents instead.

Availability and Requirements

Agent view is available today as a Research Preview on Pro, Max, Team, Enterprise, and Claude API plans. Normal rate limits still apply across all tiers.

The feature requires Claude Code v2.1.139 or later, and administrators can disable it for an organization through managed settings. Opting in is as simple as running claude agents from any terminal.

What This Means for the Industry

For Anthropic, this release fits its broader push to make Claude Code less like a chat-driven coding tool and more like an agent operations layer for software teams. The company has already been building around subagents, agent teams, skills, hooks, Remote Control, scheduled prompts, and Claude Code on the web.

At Anthropic, the majority of code is now written by Claude Code. Engineers focus on architecture, product thinking, and continuous orchestration: managing multiple agents in parallel, giving direction, and making the decisions that shape what gets built.

Rakuten reportedly reduced the average delivery time for new features from 24 working days to 5. Engineers now run multiple Claude Code sessions in parallel, delegating tasks across the codebase simultaneously. Agent view is the interface layer that makes that kind of parallel delegation practical at scale.

The feature sits alongside related tools including subagents for reusable agent configurations with custom prompts and tools, agent teams for coordinating multiple sessions that message each other, and Claude Code on the web for running sessions in a managed cloud environment.

Final Thoughts

Agent view is a focused, practical addition. It doesn't change what Claude Code can do technically, but it changes how much of it you can actually run at once without losing track. The /bg flag and the peek panel are the two interactions I'd watch most closely. Replying inline without attaching to a full session is exactly the kind of low-friction interrupt that keeps parallel work from becoming a context-switching nightmare.

The research preview label means the keyboard shortcuts and interface layout could shift. If you're building team workflows around specific key bindings right now, keep that in mind. The disableAgentView admin toggle also signals that Anthropic is thinking about enterprise deployment constraints from day one, which is a sensible call.

If you're already running Claude Code for multi-step work, update to v2.1.139 and run claude agents. It takes about 30 seconds to get oriented. What do you think of the new agent view? Drop your thoughts in the comments.

---