Anthropic has debuted a new capability for its Claude Code platform, called Channels, that enables a number of key improvements. Notably, it allows users to communicate with the coding assistant through messaging platforms like Telegram and Discord. More important, Channels is continuously available, so it is closer to an active collaborator.
Clearly, Anthropic’s release of Channels is competitive move to keep pace with OpenClaw, an open-source project that has gained massive traction since its launch in late 2025. OpenClaw popularized the concept of persistent AI agents, systems that remain active and responsive, capable of receiving instructions and completing tasks asynchronously through familiar chat applications. With Channels, Anthropic is adopting a similar model while emphasizing tighter controls and a more structured user experience.
Using MCP
With Channels, developers can send instructions to Claude Code from a mobile device and receive updates when tasks, such as debugging or code compilation, are completed. This reduces reliance on fixed workstations and allows for a more continuous workflow. The change also moves from traditional prompt-and-response interactions toward a more ongoing relationship between user and system.
Underlying this shift is Anthropic’s MCP, an open standard designed to connect AI models with external tools and data sources. Within the Channels framework, MCP acts as a bridge, enabling messages from external platforms to feed directly into an active coding session. Once a request is received, Claude Code performs tasks and returns results through the same messaging channel.
This architecture enables persistence, a feature that has been central to OpenClaw’s appeal. Instead of requiring users to initiate each session manually, the system can remain active in the background, waiting for new instructions. For developers, this means the AI assistant functions less like a tool and more like a continuous collaborator.
Anthropic’s approach also addresses concerns about autonomous agents. OpenClaw and similar tools often require extensive access to local systems, raising worries about data security and system integrity. In some regions, these concerns have prompted restrictions on their use in government or business environments. Anthropic, by contrast, is positioning Claude Code as a more controlled alternative, with admin oversight and safeguards like approved communication channels and restricted sender lists.
Intense Competition
The competitive landscape is evolving faster than you can say AI agent. OpenClaw’s open-source model has led to a swarm of variants, including projects aimed at improving usability and reducing risk. At the same time, major tech firms are exploring similar agent-based systems, so persistent AI assistants will become a standard interface for software development.
Anthropic’s release also highlights a major trend: the blending of proprietary AI systems with open standards. While Claude Code itself remains a commercial product, the use of MCP allows external developers to build integrations and extend functionality. This hybrid approach enables ecosystem growth without giving up control over the core technology.
Early reactions from the developer community suggest that the update could reduce the need for complex, self-hosted solutions that have been required to run persistent agents. By integrating these capabilities directly into a managed platform, Anthropic is lowering the barrier to entry for less technical users while retaining appeal for experienced developers.