Anthropic advises folding subagents into one Claude agent loop
Anthropic published guidance on October 22, 2026 on when to fold subagents into a main agent loop when building agents in Claude Platform, along with a migration kit that runs a guided workflow to identify patterns, evaluate alternatives and verify the best approach, including a demo that can be tried on a mock system first. The company says its Applied AI team has seen enterprise deployments benefit from folding routers, planners and domain subagents into the main loop, and that in its own comparisons a single agent with skills frequently outperformed both a one-prompt design and a subagent design on quality, often at lower cost and latency per task. Anthropic adds that subagents remain useful for verification, parallel work and modular permissioning, and that it uses them in its own agents built on Claude Platform, such as its multi-agent research system.
Anthropic is publicly recommending simpler single-loop agent designs for enterprises, but the quality, cost and latency comparisons behind that advice are its own internal results and were not independently measured.
If you’ve built an agent in Claude Platform that includes multiple subagents, see which may make sense to fold into the main agent loop using our migration kit. It runs a guided workflow to identify your patterns, evaluate alternatives, and verify the best approach. It includes a demo that you can try on a mock system before running on your agent.
Subagents – agents with their own context window, tools, and instructions, orchestrated by a coordinator – are a core component of agent design. They are effective at verification, parallel work, and modular permissioning. We use them in many of our own agents that we build using Claude Platform, for example our multi-agent research system.
However, recently the Applied AI team has found a pattern of enterprise agent deployments that have benefited from folding routers, planners, and domain subagents into the main agent loop. In our comparisons across several enterprise deployments, a single agent with skills frequently outperformed both the one-prompt-for-everything design and the subagent design on quality, and often at a lower cost and latency per task.