The revamped projects feature in Claude Code now enables users to manage a team of AI agents in the cloud, with shared memory and goals, much like running multiple threads in parallel with a coordinator overseeing everything.

Each thread functions as a Claude Code cloud session, branching off from the main repository, and can be further divided into subagents, loops, and workflows for more complex tasks. Users can interact with each thread individually or monitor the project as a whole through the main chat interface.

Anthropic, the company behind Claude Code, has started the beta phase for select Claude Pro and Max subscribers, with broader access planned for other users over the coming months. While the initial rollout is cloud-based, the company has announced support for local tools and code is imminent.

This update marks a significant step in making AI tools more collaborative and efficient, potentially revolutionizing how teams work with AI agents. Whether this leads to smarter offices or just more AI-led procrastination, only time will tell.