Anthropic is redesigning Claude projects so users can describe a broad task in one conversation while Claude organizes work across multiple parallel threads. Anthropic announces in an official Claude blog post that the updated Projects feature begins rolling out in beta for selected Claude Pro and Max subscribers who use Claude Code cloud sessions.
The new system centers on a coordinator that plans and assigns work, plus individual threads that complete the assigned tasks. A user could, for example, set a goal to improve an app’s checkout speed. Claude can then create separate threads to examine endpoints, test changes, and prepare pull requests.
Each thread runs as its own Claude Code cloud session, with a separate copy of the code repository and its own branch. This allows work to proceed simultaneously, but it does not eliminate standard software development issues. If two threads change the same code, their work can still create a merge conflict that must be resolved.
Shared memory for longer projects
Anthropic says Projects are intended for work that extends beyond a single response. Threads contribute to a shared memory, which can retain project decisions, deadlines, dropped ideas, and preferences about communication. Projects also include a library for uploaded files and materials created during the work.
Users can follow progress in a central project chat or open individual threads to give more detailed direction. They can select the model and effort level for both the coordinating chat and worker threads, and check project-specific usage.
The beta currently excludes people who already have Claude projects on the web or desktop. Anthropic says it plans to extend access to more eligible Claude Code users, then to its wider products and Team and Enterprise plans. Threads currently run in the cloud. Support for work on local machines and within private networks is planned for a later release.
Video
Stay up to date
AI for content creation: the latest tools, tips and trends. Every two weeks in your inbox:
