Claude Code projects get a coordinator that runs parallel threads in beta

Claude Code projects get a coordinator that runs parallel threads in beta

A claude.com article announces a redesign of Claude projects, now in beta inside Claude Code. The pitch: managing several sessions across a build used to mean dividing the work, juggling handoffs and stitching results together by hand. In the new setup, you describe what needs to get done and Claude manages it. Claude scopes the request, delegates the work, coordinates parallel threads, reviews the outputs and assembles the finished result. You can steer progress throughout, even from your phone, and the work continues after you step away from your computer.

The article gives two examples. In the first, you configure a project and set a goal to reduce your app's checkout p75 latency, then ask Claude to profile each endpoint, test optimizations and open PRs in parallel threads. In the second, you connect your API, web and mobile repos and set a goal to retire a deprecated v1 endpoint. Claude creates a thread per repo to migrate the callers, run the tests and open PRs, then tells you which ones need to merge first. Both are illustrations, not measured results.

The structure is simple: projects have threads that do the work and a coordinator that directs them. When you start a project, you pick a goal as well as the repo or context. Claude begins by suggesting work it can pick up right away. You can configure the project's cloud environment, connectors, plugins, instructions and model. Progress can be monitored and guided in the main project chat, or you can open any single thread to examine and steer the details. The article says to brief Claude the way you would brief a chief of staff, and it routes work to new or existing threads. Claude also checks in and follows through: with repositories connected, a thread opens pull requests and runs your tests; with documents, it reads them and drafts.

Under the hood, each thread is a Claude Code cloud session working on its own branch and copy of the repo. The coordinator keeps work organized, but if threads touch the same code, the overlap is resolved as a merge conflict, like any other PR. Each thread can also split its delegated work further using subagents, loops and workflows, so large assignments finish faster.

Projects are meant for long-running or agentic work, meaning tasks that take longer than one reply and have more than one part. Every thread adds to and draws from a shared project memory, which the article says reduces the need for complex prompt engineering. Examples of what Claude can remember: that the release moved to Friday, why the export was dropped, or who to check with before touching the billing service. It also remembers your working and communication style, and you can ask it to change how often it checks in, how often it starts new threads, or how detailed each update is. Projects now also include a library that collects the files you add and the artifacts Claude produces, so new work can build on past efforts.

Availability: starting today, updated projects are in beta for select Claude Pro and Max subscribers who use cloud sessions in Claude Code and have no existing projects on the web or desktop. Over the coming week, access expands to more Claude Code users on those plans. Updated projects across all of Claude and on Team and Enterprise plans come after that. Pro and Max users without access can join a waitlist. Existing projects on Pro and Max plans keep working as they do today and will be upgraded as the rollout expands to chat and Cowork.

The article closes with two caveats under "What's next". Because a project can run several threads at once and each is a full Claude Code session, projects can reach usage limits faster; you can check project-specific usage and choose the model and effort levels for the coordinator chat and the worker threads. And threads run in the cloud today; running on your own machine, alongside your local tools and code and behind your network, is coming very soon.

Key facts

  • Claude projects in Claude Code now have a coordinator that scopes a goal, delegates to parallel threads, reviews outputs and assembles the result.
  • Each thread is a Claude Code cloud session with its own branch and copy of the repo; overlapping edits surface as ordinary merge conflicts.
  • All threads in a project read from and write to a shared memory, and a new library collects added files and Claude-made artifacts.
  • The beta starts with select Pro and Max subscribers who use cloud sessions and have no existing projects; access widens over the coming week, with all of Claude and Team and Enterprise after that.
  • Several full sessions at once means projects can hit usage limits faster; threads run only in the cloud for now, with local running described as coming very soon.

Why it matters

This moves Claude Code from a single session you drive toward a setup where one coordinator drives many sessions. The work the article says users used to do by hand, dividing tasks, handing off between sessions and stitching results back together, is handed to Claude. The examples are concrete: profile every endpoint and open PRs in parallel for a latency goal, or migrate callers across API, web and mobile repos and report which PRs must merge first. Shared memory and a file library are meant to let context build across threads over time.

Who it affects

First in line are select Claude Pro and Max subscribers who use cloud sessions in Claude Code and have no existing projects on the web or desktop. Access is due to widen to more Claude Code users on those plans over the coming week. Team and Enterprise customers, and projects across all of Claude, come after that, with no dates given. People with existing projects on Pro and Max see no change for now; those projects will be upgraded as the rollout reaches chat and Cowork.

How to use it

Start a project, choose a goal plus the repo or context, and configure the cloud environment, connectors, plugins, instructions and model. Claude will suggest work it can start on at once. Brief it in the main project chat as you would a chief of staff, and it routes work to new or existing threads; open any thread to inspect or steer it. You can ask Claude to adjust how often it checks in, how often it starts threads and how detailed its updates are. Pro and Max users without access can join the waitlist. You can check project-specific usage and pick the model and effort levels for the coordinator and for the worker threads. The article gives no pricing.

How solid is it

The source is a first-party article on claude.com announcing a beta. The examples, such as the checkout p75 latency goal and the v1 endpoint retirement, are illustrations. No benchmark or measured speed-up is reported, no outside reactions or independent testing are included, and no figure is given for how many threads can run in parallel.

Risks and caveats

Usage is the stated cost: several threads at once, each a full Claude Code session, means projects can reach usage limits faster, and no limit figures are given. Threads run only in the cloud today; running on your machine, with local tools and behind your network, is described only as coming very soon, with no further timeline. If threads work on the same code, you still face a merge conflict like any other PR. The rollout is staged and limited to select users at first.

“Projects have threads that do the work and a coordinator that directs them.”

— claude.com article, "Projects redesigned: from folder to conversation"