Anthropic has updated the project section of Claude Code to become a workspace for multiple AI agents. They can use shared memory, targets, files, and artifacts to perform different tasks simultaneously in the same project.
Image source: Human
Separate threads are created within the project, each thread performs its own task, and they are coordinated by a special coordination agent, which distributes the work and ensures that the tasks are completed in an organized manner.
Each thread represents a separate Claude Code cloud session, with its own branch and copy of the repository. If multiple agents modify the same code, the intersection will be treated as a merge conflict, just like when dealing with regular pull requests (PR). If necessary, individual threads can further delegate their tasks to subagents, loops, and worker threads.
Users can interact with each thread individually or monitor all work through the project’s main chat to clarify edge. At launch, the agent will only run in the cloud, but Anthropic promises to add support for local tools and code “soon.” The updated workspace is already available in beta to select Claude Pro and Max subscribers, and will be rolled out later to all Pro, Max, Team and Enterprise users, as well as Cowork and regular Claude chat rooms.
If you find an error, select it with your mouse and press CTRL+ENTER.