Anthropic Launches Claude Code Projects, an Always-On Conversation that Remembers and Delegates Your Long-Running Dev Work
VentureBeat, Thursday, September 17th, 2026
VentureBeat reports that Anthropic's Claude Code Projects adds a persistent, always-on coordination layer for long-running dev work.
VentureBeat covers Anthropic's launch of Claude Code Projects, a persistent coordination layer that keeps a single always-on Claude conversation running across development sessions.
The agent automatically subdivides a developer's request into parallel worker threads and maintains shared memory and project context between them, so decisions and dependencies survive without manual context re-establishment.
Execution is cloud-based at launch, with local execution promised soon. VentureBeat frames the release against rival agent-orchestration control planes from OpenAI, GitHub and Google, each of which places the coordination boundary in a different place.