Can Claude Code sub-agents call each other?

The main issue is how far Claude Code can split work between agents. The current understanding is that the main agent gives a specific task to a sub-agent, the sub-agent uses tools if needed, and then returns the result to the main agent.

The open question is whether one sub-agent can call another sub-agent directly, and whether several sub-agents can talk to each other without always going through the main agent. A broader goal is a where agents collaborate, coordinate, review, or negotiate with each other.

The software engineering example has separate roles: a Planner Agent creates a plan, a improves the , a implements it, a Agent checks the result, and a Debug Agent investigates failures. The desired setup is a multi-level system where agents can call each other when needed and exchange context directly.

Key points

  • Claude Code is understood here as using a main agent that delegates tasks to sub-agents.
  • The key question is whether a sub-agent can directly call another sub-agent.
  • Direct communication between multiple sub-agents is also in question.
  • The intended use case is a software workflow split into planning, design, coding, , and debugging roles.
  • The main thing to verify is whether Claude Code supports this kind of multi-level .
Read original