Anthropic has officially relaunched its Projects feature in Claude Code to enable multi-agent team management within cloud development environments. The updated functionality allows developers to run multiple AI agents under a single project structure equipped with shared memory, common objectives, and a central repository for files and artifacts.
Under this architecture, a main coordinator agent divides assignments among multiple parallel sub-threads, each running in a dedicated cloud session on its own git branch. Each sub-thread can spawn additional subagents, loops, and workflows to complete delegated tasks. When parallel threads edit identical files, overlaps are managed and resolved as standard code merge conflicts.
Currently available in beta for select Claude Pro and Max users, Anthropic plans to roll out the updated Projects feature across all Pro, Max, Team, and Enterprise plans, with local tool execution support slated for release in the near future.
Why it matters
Enables software teams to manage parallel AI subagents working concurrently on distinct feature branches within a single project.
Streamlines enterprise codebase management by resolving agent-generated code overlaps through standard git merge processes.
Signals a commercial push toward cloud-hosted agent environments ahead of upcoming local tool integration.
Source: theverge.com
