Anthropic redesigns Claude projects, letting users describe work in one conversation and have Claude manage it across parallel threads, starting in Claude Code

Your coding assistant can now manage complex, multi-stage development tasks autonomously and continuously.

Anthropic has launched a significant update to its Claude Code projects, enabling users to define a task in a single conversation and have Claude manage its execution across multiple, parallel threads. This new functionality allows Claude to scope requests, delegate work, coordinate concurrent threads, review outputs, and assemble final results. Users can provide direction throughout the process, even when away from their computers. For example, a user could instruct Claude to optimize an app's checkout latency by profiling endpoints, testing optimizations, and opening pull requests in parallel. Another use case involves retiring a deprecated API endpoint by having Claude create separate threads for each repository to migrate callers, run tests, and open pull requests, then indicating the merge order. The feature is rolling out gradually, starting with select Claude Pro and Max subscribers using cloud sessions in Claude Code, with wider access planned for more users and eventually all Claude plans.

This redesign transforms AI coding assistants from tools that execute discrete commands into autonomous project managers capable of handling complex, long-running workflows. By managing parallel threads, resolving conflicts, and maintaining project context over time, Claude Code projects are poised to significantly accelerate software development cycles and reduce the burden of intricate coordination for development teams. The system's ability to learn and adapt to user preferences and project specifics, coupled with a shared memory across threads, streamlines iterative development and complex problem-solving.

The implications extend to how software projects are scoped and executed, potentially democratizing advanced development tasks and enabling smaller teams or individual developers to tackle more ambitious projects. As this functionality expands beyond Claude Code to broader Claude applications, it signals a market shift towards more integrated, agentic AI solutions that can operate semi-independently. Developers should anticipate a future where AI handles substantial portions of the development lifecycle, freeing human expertise for higher-level strategy and innovation, and users will need to adapt to guiding and reviewing AI-driven project execution.