Claude Code experimental agent teams is here
Claude Code now has an experimental option to enable a team of agents, and here is how it's explained by Lydia Hallie:
Instead of a single agent working through a task sequentially, a lead agent can delegate to multiple teammates that work in parallel to research, debug, and build while coordinating with each other.
It's disabled by default but to enable this, you need to add the following in your Claude's settings.json file.
// add this in `settings.json` file
{
"env": {
"CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS": "1"
}
}You can learn more about this agent teams feature from this page in their docs.
Subagents only report results back to the main agent and never talk to each other. In agent teams, teammates share a task list, claim work, and communicate directly with each other.
By the way, while Subagents and Agent Teams both let you parallelize work, their working mechanism is different as demonstrated above (from Anthropic docs page). There's a lot more info in their documentation about this that you can further explore.
And to actually trigger an agent team, you need to mention something like "create an agent team" or something similar.
I also saw a post about the same from Ian Nuttall saying how Claude Code TeamCreate/swarm feature is so good. And also came across a bunch of other people trying the feature, but I am yet to try this.
Webmentions