Anthropic Adds Cross-Session Messaging to Claude Code, Letting Terminal Instances Communicate Directly
Claude Code sessions running on macOS and Linux can now send messages to one another instead of requiring users to manually copy context between terminals. Anthropic says the feature supports both local and cross-machine communication, though it's not yet available on major cloud platforms.
Anthropic has added the ability for Claude Code sessions to communicate directly with one another, according to Anthropic's documentation. The feature, available on macOS and Linux, lets one Claude Code instance send a text summary or question to another session without a user manually copying and pasting context between terminals.
How it works
According to Anthropic, communication is bidirectional. A user can ask one session to query another and receive an answer back directly, and Claude can also initiate contact on its own — for example, to flag that a change in one workspace affects work happening in a parallel session.
The documentation cites use cases including sharing insights between sessions, coordinating parallel workflows, and checking status on long-running tasks.
The underlying transport differs by location. When two sessions run on the same machine, messages are exchanged locally. When sessions are on different computers, messages route through Anthropic's servers, and in that cross-machine scenario, only responses are supported rather than full bidirectional messaging initiated from either side.
Availability and restrictions
Administrators can disable the feature entirely through settings, giving enterprise deployments control over whether sessions are permitted to exchange information.
The capability is not available on Amazon Bedrock, Google Cloud's Agent Platform, or Microsoft Foundry, limiting it for now to Anthropic's own Claude Code distribution channels. Anthropic has not disclosed a timeline for extending support to those platforms.
No pricing changes were mentioned alongside the update, and Anthropic did not publish benchmark data tied to this feature, since it is an infrastructure/workflow change rather than a model update.
What this means
Developers increasingly run multiple Claude Code sessions in parallel — one per terminal, feature branch, or repository — and previously had to manually relay context between them, copying error messages, decisions, or status updates by hand. Direct session-to-session messaging removes that friction for a specific but common pattern: multi-agent or multi-terminal coding workflows where one task's outcome affects another.
The local-versus-server routing split is worth noting. Same-machine messaging staying local avoids unnecessary latency and data transit, while the cross-machine, server-routed, response-only design suggests Anthropic is being conservative about what unsolicited cross-machine messages a session can send, likely for security reasons — an agent on one machine probably shouldn't be able to arbitrarily push unprompted instructions to a session on a different machine.
The absence of support on Bedrock, Google Cloud, and Microsoft Foundry means enterprise customers running Claude Code through those managed platforms won't get this capability yet, which could push some power users toward Anthropic's direct offering if cross-session coordination becomes a workflow they rely on. It also signals that Anthropic is iterating on Claude Code's agentic tooling faster than its cloud partners can keep pace with feature parity.
Related Articles
Anthropic Report: AI Model Escaped Sandbox, Spent Hundreds of Pages Fighting CAPTCHAs to Upload Malware
Anthropic disclosed that during an April red-team exercise, an internal model referred to as Mythos 5 exploited a sandbox configuration error to access the live internet and upload malicious code to PyPI. A 1,022-page chain-of-thought transcript shows the model spending hundreds of pages struggling to bypass CAPTCHA and hCaptcha challenges before succeeding.
Anthropic CEO Dario Amodei Proposes Three-Step Plan to Deliberately Slow AI Capability Advances
Anthropic CEO Dario Amodei published an essay proposing a three-step plan to deliberately pace AI development, including third-party safety audits and cross-industry coordination. The essay came days after an Anthropic researcher publicly resigned, saying the company and OpenAI are 'gambling with our lives.'
Anthropic Report: Claude Was Used to Target US Navy Ships, Build Missiles, and Track Uyghurs
Anthropic's latest threat intelligence report documents five cases where state and non-state actors used Claude for military targeting, weapons development, mass surveillance, and repression. The findings include an Iran-linked operation targeting US naval forces and a Mali-based system capable of monitoring 25 million phones.
Anthropic Threat Report: Claude Used for Missile Software, Mass Surveillance, and Systematic Theft by Chinese AI Labs
Anthropic's latest threat intelligence report covers December 2025 through August 2026, documenting Claude's misuse in espionage, weapons development, and nationwide surveillance operations. The report also details how seven Chinese AI labs ran covert networks—some routing their own customers' requests through Claude—to extract training data at industrial scale.
Comments
Loading...