AI Pulse · AI Trends Pulse

Anthropic just gave Claude Code the ability to talk to itself. Version 2.1.224, released August 7, added SendMessage and ListAgents tools so separate Claude sessions can swap notes without you copying and pasting between windows. If you’re running one session in a terminal, another in a different project, and a third on another machine, they can now coordinate directly.

This builds on Agent Teams, a feature that already let Claude instances share a task list and a mailbox. The new tools make that coordination automatic. One session can ask another what it’s working on, hand off a subtask, or flag a blocker, all while you’re doing something else. The documentation frames it as a way to keep context flowing when you’re juggling multiple workstreams.

Why it matters

Most businesses hit the same wall with AI assistants. You ask one to draft a report, another to pull data, a third to check compliance. Then you spend half your morning shuttling information between them. This update removes that step. The agents handle the handoffs.

It’s the same pattern we built into EDNA’s swarm architecture, where specialist agents dispatch tasks and share context without a human in the loop. If you’re running operations that involve multiple AI tools or workflows, the ability to let them coordinate quietly in the background saves real time. It’s not flashy, but it’s the kind of plumbing that turns AI from a set of separate tools into something that works more like a team. We design exactly this kind of orchestration into the Omni Command Centre, so different AI capabilities can hand off work without you managing every step.

If you’re already using Claude Code across projects, the update is live. If you’re not, it’s worth watching how agent-to-agent messaging changes what’s practical to automate.