Claude Code has launched a revamped Projects feature that enables users to manage and run multiple AI agents simultaneously within a single environment. The system provides shared memory, goals, and resources across agents, operating similarly to existing multi-agent management tools, and is significant because it allows users to orchestrate complex, parallelised work across multiple agent threads more efficiently.

The Projects system organises work through "threads," each running as an independent Claude Code cloud session on its own branch of a repository. A coordinator oversees the parallel threads and resolves any code conflicts as merge requests would be handled in standard version control. Individual threads can further delegate tasks using subagents, loops, and workflows to accelerate completion. The feature is initially available in beta to select Claude Pro and Max subscribers, with rollout planned to all Pro, Max, Team, and Enterprise users, as well as Cowork and regular Claude chats. Whilst threads currently run in the cloud, Anthropic has indicated that local tooling and code support will arrive "very soon."