Anthropic adds always-on channels to Claude Code, enabling async AI agent capabilities
Anthropic has added "channels" to Claude Code, enabling Claude to respond to incoming messages, webhooks, and notifications asynchronously without user intervention. The research preview supports Telegram and Discord with custom channel support, running through MCP servers with two-way communication.
Anthropic adds always-on channels to Claude Code, enabling async AI agent capabilities
Anthropric has launched a channels feature for Claude Code that transforms the tool into an always-on AI agent capable of responding to events without user presence at the terminal.
What's new
Channels allow messages, notifications, and webhooks to flow directly into a running Claude Code session. The feature enables two-way communication: Claude reads incoming messages and responds through the same channel. This means Claude can now react to CI/CD results, chat messages, and monitoring alerts in real time, whether the user is actively using the terminal or not.
The implementation runs through MCP (Model Context Protocol) servers, maintaining Anthropic's strategy of building extensibility into its tooling infrastructure.
Current availability and requirements
The feature is in research preview status. Supported integrations include:
- Telegram (official support)
- Discord (official support)
- Custom channels (developer-buildable)
Technical requirements:
- Claude Code version 2.1.80 or later
- claude.ai login (API key authentication not yet supported)
- Teams and Enterprise organizations must explicitly enable channels through settings
The API key limitation suggests this feature is currently bound to web-authenticated sessions rather than programmatic workflows, potentially constraining enterprise automation use cases.
Strategic positioning
This update positions Claude Code closer to the autonomous agent narrative that has dominated AI development discourse, particularly following competitive releases. The async event-driven model differs from persistent agent architectures—Claude responds to incoming stimuli rather than operating continuously without external triggers, but the effect is functionally similar for many use cases.
The research preview status indicates Anthropic is testing community interest and real-world usage patterns before broader release. Pricing for channels and any future API-native implementation has not been disclosed.
What this means
Channels expand Claude Code's utility beyond interactive development sessions into workflow automation, monitoring, and distributed team collaboration. The restriction to claude.ai authentication suggests Anthropic is prioritizing stability and safety before opening this to programmatic API access. Developers building agent workflows will need to evaluate whether event-driven response patterns meet their requirements compared to continuously-running agent architectures. The custom channel support creates a pathway for integration into existing DevOps and communication stacks without waiting for official Anthropic integrations.
Related Articles
METR report: Anthropic's unreleased Model 4 scores 178.6 on AECI, ahead of Opus 5.5
METR's new report discloses three unreleased Anthropic models, Model 2, Model 3 and Model 4. Model 4 scores 178.6 on Anthropic's AECI index, about 3.9 points above Opus 5.5. Anthropic co-founder Sam McCandlish confirmed the report's main findings but said much information was removed.
METR Report: Anthropic's AI Development Pace May Have Doubled Due to Internal AI Use
A new report from AI research nonprofit METR says AI use within Anthropic may have doubled the company's pace of AI development. The rate at which Anthropic's models are improving as of October is roughly twice the rate between March 2024 and October 2025, according to The Information.
Anthropic launches Personal Finance in Claude with read-only Plaid access for US Pro, Max and Ultra users
Anthropic announced Personal Finance in Claude, a feature that connects bank and credit card accounts through Plaid with read-only access. Conversations in the Personal Finance hub are not used to train Anthropic's models. It is available only to Pro, Max and Ultra subscribers in the US.
Moonshot AI launches Kimi Hosted Agents, a managed agent runtime with sandboxes at 0.072 yuan per hour
Moonshot AI has launched Kimi Hosted Agents, a Harness as a Service product that runs the agent runtime on Kimi's infrastructure. It offers isolated sandboxes, exportable audit logs, more than 140 skills and more than 30 plugins. IT Home cites a sandbox rate of 0.072 yuan per hour plus model and plugin charges.
Comments
Loading...