Anthropic is developing a new tool called Claude Conway that includes features like scheduled triggers and the ability to persist across container restarts. This initiative is part of Anthropic’s broader strategy to enhance agentic workflows, allowing Conway to remain active and responsive to external events beyond single sessions. The new tool is positioned to compete with other agent platforms by incorporating a native extension layer and deeper integration with Claude Code, indicating a significant evolution in Anthropic’s product offerings.

Conway: Conway is an upcoming always-on agent tool and standalone environment being built by Anthropic for its Claude models. It is designed to support persistent instances that survive restarts, handle recurrent scheduled triggers, external webhooks, Chrome integration, notifications, and a custom extension framework including .cnw.zip packages. The tool introduces dedicated sidebar sections for Search, Chat, and System management, along with connectors, tools, and potential operator interfaces such as Epitaxy.
Anthropic: Anthropic is an artificial intelligence research company known for building the Claude family of large language models with a focus on helpful, honest, and harmless AI systems. The company is actively expanding its product lineup beyond standard chat interfaces toward more advanced agentic capabilities. In this news, Anthropic is developing Conway as a dedicated always-on agent environment that integrates deeply with Claude Code and external systems.

Agentic Workflows: Anthropic is advancing its product direction around Claude Code by building persistent agent environments that can respond to external events beyond single sessions.
Platform Comparisons: The Conway development moves Anthropic closer to other agent platforms while emphasizing native extension layers and deeper Claude integration.