Anthropic redesigns Claude Code Projects with AI orchestration, cloud-only rollout
New features let developers coordinate multiple threads and share memory across repos, but local development remains unsupported.
What to know
- Claude Code Projects now include AI orchestration that coordinates parallel threads, manages shared memory, and spans multiple repositories without manual intervention.
- The redesigned feature is limited to select Claude Pro and Max cloud subscribers; local developers cannot use the new capabilities and must wait for future rollouts.
- Key features include mid-flight steering of named threads, shared memory that reduces prompt engineering complexity, and automatic work delegation across repos.
“Over time, Claude learns more about the project details and applies them to its work. Every thread now adds to and draws from a shared memory, reducing the need for complex prompt engineering.”
Anthropic, AI company · ZDNet ↗ · Sep 16
Anthropic AI company behind ClaudeDavid Gewirtz Technology analyst and columnist
How it unfolded 1 development · click the chart to see its coverage articles
-
1
Feature limited to cloud subscribers; local developers excluded from initial rollout
The redesigned projects are available only to select Claude Pro and Max subscribers using cloud sessions who don't have existing projects. Local development environments are not supported in this initial release, and the company stated features will roll out to other users over time.
“Claude scopes the request, delegates the work, coordinates parallel threads, reviews the outputs, and assembles the finished result. You can steer progress throughout, and it keeps working after you close your laptop.”
— Anthropic, AI company · source -
background
Anthropic announces Claude Code Projects redesign with orchestration and threading — Anthropic introduced a revamped Claude Code Projects feature that enables AI-driven orchestration of parallel threads, shared memory across threads, and multi-repository workflows. The coordinator scopes requests, delegates work, coordinates threads, and reviews outputs without requiring manual intervention between sessions.