AI

Anthropic rebuilds Claude Code desktop app around parallel sessions

At a glance:

  • New sidebar organizes active and recent sessions, filterable by project or environment
  • Integrated terminal, file editor, diff viewer, and preview pane now live in the desktop app
  • Routines automate tasks without an active session, running on Claude Code's web infrastructure

Parallel sessions and a new sidebar

Anthropic has overhauled the Claude desktop app with a focus on managing multiple coding sessions at once. The centerpiece is a redesigned sidebar that lists every active and recent session in a single view. Users can filter sessions by status, project, or environment, and group them by project for easier navigation. A new keyboard shortcut—Command + ;—lets developers branch a side chat from a running task without polluting the main thread with extra context.

Built-in developer tools

The update brings several developer workflows directly into the desktop interface. An integrated terminal allows running tests and builds without leaving the app. An in-app file editor supports quick edits, while a rebuilt diff viewer is optimized for large changesets. The expanded preview pane can render HTML files, PDFs, and local app servers. Each pane is drag-and-drop friendly, so layouts can be rearranged to suit individual workflows.

Plugin support, SSH, and view modes

Plugin support now matches the Claude Code CLI, giving desktop users access to the same extensibility. SSH sessions are supported on both Mac and Linux, broadening remote development options. Three view modes—Verbose, Normal, and Summary—let users control how much of Claude's tool-call activity appears on screen, from detailed logs to high-level summaries.

Routines: automations without an active session

Alongside the desktop refresh, Anthropic introduced Routines, a new automation feature for Claude Code. A routine bundles a prompt, a repository, and any relevant connectors into a single configuration that can run on a schedule, fire from an API call, or trigger off a GitHub event such as a new pull request. Routines execute on Claude Code's web infrastructure rather than a local machine, with daily run caps that scale by subscription plan. The feature is available in research preview to Pro, Max, Team, and Enterprise users who have Claude Code on the web enabled.

Availability

The desktop app update is rolling out now to Claude Code users on Pro, Max, Team, and Enterprise plans. Routines are also in research preview for the same tiers, provided Claude Code on the web is activated.

Looking ahead

By unifying session management, embedding core developer tools, and introducing cloud-based automations, Anthropic is positioning Claude Code as a more autonomous coding partner. The emphasis on parallel sessions and Routines suggests a longer-term vision where Claude can handle multiple tasks—both interactive and unattended—across a developer's workflow.

Editorial SiliconFeed is an automated feed: facts are checked against sources; copy is normalized and lightly edited for readers.

FAQ

What are the main new features in the Claude desktop app update?
The update introduces a sidebar for managing active and recent sessions, an integrated terminal, in-app file editor, rebuilt diff viewer, expanded preview pane for HTML and PDFs, drag-and-drop layout, SSH support on Mac and Linux, plugin support matching the CLI, and three view modes (Verbose, Normal, Summary).
What are Routines and how do they work?
Routines are cloud-based automations that run without an active session. They bundle a prompt, repository, and connectors into a configuration that can execute on a schedule, via API, or in response to GitHub events like new pull requests. They run on Claude Code's web infrastructure with daily caps that scale by plan.
Who can access the new Claude desktop app features and Routines?
Both the desktop app update and Routines are available in research preview to Pro, Max, Team, and Enterprise users who have Claude Code on the web enabled.

More in the feed

Prepared by the editorial stack from public data and external sources.

Original article