Anthropic has relaunched its Projects feature in Claude Code to support the coordination of multiple AI agents within a single cloud workspace. The update enables users to run several autonomous agents simultaneously, each operating in its own branch of a code repository while sharing memory, objectives, and access to files and artifacts. A central coordinator manages task allocation across these agents, referred to as threads, and resolves conflicts that arise when multiple threads modify the same code segment, mirroring the merge conflict handling used in standard pull requests. Individual threads can further delegate work through subagents, loops, or workflows to accelerate large-scale tasks. Users interact with each thread directly or via the primary project chat interface, and Anthropic plans to introduce support for local tooling and code execution in the near future. The beta version is currently accessible to a limited group of Claude Pro and Max subscribers, with broader rollout scheduled for all Pro, Max, Team, Enterprise, Cowork, and standard Claude users. The feature is designed to streamline complex development workflows by distributing workloads across parallel AI processes while maintaining centralized oversight.