Claude Code Projects Returns with Cloud-Based Agent Teams

Claude Code Projects Returns with Cloud-Based Agent Teams | AI News GlobAI

Short answer: On September 17, 2026, Anthropic relaunched Claude Code Projects, letting users run a group of AI agents together in a cloud environment with shared memory, common goals, and a single library. Each stream is an independent Claude Code session overseen by a coordinator that flags code overlaps as merge conflicts and supports subagents or reusable workflows. The feature is in beta for Claude Pro and Max subscribers, with broader access planned soon.

Claude Code Projects relaunched with cloud-based AI agent teams

On September 17, 2026, Anthropic announced that the Projects feature inside Claude Code has been relaunched, now letting users run a group of AI agents together in a cloud environment. The update gives each project a shared memory space, a common set of goals, and a single library where files and artifacts live for all agents to read and write. This design mirrors other systems that orchestrate multiple AI workers, such as Grok Bot, by splitting work into parallel streams that are overseen by a central coordinator.

How Claude Code Projects streams work as independent cloud sessions

Inside each project, every stream operates as an independent Claude Code session hosted in the cloud, complete with its own branch and copy of the repository. The coordinator watches over the streams, making sure their efforts stay aligned. When two streams happen to edit the same piece of code, the system treats the overlap like a merge conflict in a pull request, automatically flagging it for resolution. Developers can also tell a stream to break its assigned work into smaller pieces using subagents, loops, or reusable workflows, which helps large tasks finish faster than a single agent could manage on its own.

Interaction with the setup is flexible. Users can open a chat with any individual stream to give it direct instructions or check its progress, or they can stay in the main project conversation to monitor overall status and adjust priorities from one place. At launch, all streams run exclusively in Anthropic’s cloud infrastructure, although the company said that support for local tools and on-premise code repositories is expected to arrive very soon.

Claude Code Projects beta release for Claude Pro and Max subscribers

The revamped Projects capability is being released in beta starting today, September 17, 2026, for a limited group of Claude Pro and Max subscribers. Anthropic plans to widen access over the coming weeks to include every Pro, Max, Team, and Enterprise customer, as well as users of the Cowork collaboration mode and standard Claude chats. No exact timeline for the broader rollout was provided, but the company indicated that the expansion will happen shortly after the beta period ends.

Benefits of Claude Code Projects for AI developers reducing manual overhead

For developers who build with AI, the change removes a lot of the manual overhead that comes with coordinating multiple language model instances. Instead of stitching together separate prompts and handling context passing by hand, teams can now rely on the built-in coordinator to keep shared state consistent and to surface conflicts early. This makes it easier to tackle ambitious code refactors, multi-module feature additions, or large-scale data pipelines without constantly switching between chats or managing ad-hoc scripts.

Readers who want to try the new functionality should first verify whether their subscription falls into the initial beta group. If they are a Claude Pro or Max user, they can check the Anthropic dashboard for an invitation to enable Projects. Those on other plans should watch for announcements about the upcoming general availability. In the meantime, it may be useful to map out a current project that would benefit from parallel agent work, outline the goals and shared files, and think about how subagents or workflows could be inserted to speed up the process. Keeping an eye on the forthcoming local-support announcement will also help teams decide whether to wait for on-premise capabilities or start experimenting in the cloud now.

Frequently asked questions

What is the Claude Code Projects feature that was relaunched on September 17, 2026?

It lets users run a group of AI agents together in a cloud environment, giving each project a shared memory space, common goals, and a single library for files and artifacts that all agents can read and write.

Who is eligible to try the Claude Code Projects beta at launch?

At launch, the beta is available to a limited group of Claude Pro and Max subscribers; Anthropic plans to expand access later to all Pro, Max, Team, and Enterprise customers, as well as Cowork collaboration mode users and standard Claude chats.

How does the system handle conflicting edits made by different agent streams?

When two streams edit the same piece of code, the system treats the overlap like a merge conflict in a pull request and automatically flags it for resolution.

What capabilities do individual streams have inside a Claude Code Project?

Each stream operates as an independent Claude Code session hosted in the cloud with its own branch and repository copy, and can be instructed directly via chat, or broken into subagents, loops, or reusable workflows to speed up large tasks.

Source: The Verge