Claude Code relaunches Projects

Anthropic's revamped Projects feature lets users run multiple AI agents under one roof with shared memory, goals, and file libraries. Agents coordinate with each other to complete complex tasks.

“Projects allow users to control a team of Claude Code agents that all coordinate with each other.”

Threads run tasks in parallel

Each project contains threads that are separate Claude Code cloud sessions working on different branches. A coordinator keeps everything organized while threads can split work further using subagents and workflows.

Merge conflicts handled like code

When multiple threads work on the same code, overlapping changes are resolved as merge conflicts just like in pull requests. This keeps coordination clean and prevents duplicate work.

Monitor through main project chat

Users can interact with individual threads or manage everything through a single project chat interface. This gives flexibility in how teams oversee multi-agent work.

Local support coming very soon

Threads currently run in the cloud, but Anthropic says local tools and code support is arriving very soon. This will give users more flexibility in where their agents operate.

Beta rolling out to Pro subscribers

The updated Projects feature launches in beta today for select Claude Pro and Max subscribers, with broader access to Pro, Max, Team, and Enterprise users coming later.