Anthropic has restructured the Claude Code project features, adding a multi-threaded parallel collaboration mechanism. Users need only describe the project goal, and the coordinator will split the task and assign it to multiple independent "threads," each running in a separate cloud session. Users can track progress in real-time through the main chat room or each thread, and mobile support is also available; threads can also independently submit pull requests and run tests.
During execution, Claude continuously builds shared memory between different threads, while the project library collects user-uploaded files and results generated by each thread, forming a collaborative workflow for complex development tasks.
Currently, the cloud session feature of Claude Code is available for testing to some Pro and Max subscribers. Team and enterprise versions will be released later, and local execution capabilities are also under development. Users can apply for experience through a waiting list.
This update further advances Claude Code from a single AI programming assistant to a multi-agent collaboration platform. Previously, Anthropic had set the autonomous driving mode of Claude Code as the default and stated that its performance in some security tasks is better than that of human developers. As more tasks shift to autonomous agents for parallel execution, the computing resource consumption, cost of use, and the way developers control agent behavior of AI programming tools will become new industry focus points.
