Hierarchical agent
A multi-agent architecture where a "manager" or "planner" agent delegates sub-tasks to specialist worker agents — the most common multi-agent pattern in 2026 production systems.
Hierarchical agents organize work the way humans organize teams. A high-level planner agent decomposes the user goal into sub-tasks. Specialist worker agents execute the sub-tasks (one might do code generation, another runs tests, another writes documentation). The planner aggregates results and delivers the final output.
In 2026 hierarchical is the dominant multi-agent pattern. It maps cleanly onto org-chart metaphors, allows specialist tuning per agent, and is debuggable (you can trace which agent did what). Devin, Manus, and most coding agents use hierarchical patterns internally.
The trade-offs: extra coordination overhead, risk of token spend on inter-agent communication. The wins: dramatically better performance on tasks that decompose cleanly into specialist sub-problems.
Frequently asked
When should I use hierarchical agents?+
When the task decomposes into independent sub-tasks that benefit from specialist tuning. Coding workflows (planner + coder + tester + reviewer), research (planner + searcher + writer), customer support (router + specialist agents per domain).
Is hierarchical the same as supervisor pattern?+
Closely related. Hierarchical describes the overall structure (manager + workers). Supervisor describes the role of the top-level agent (oversees + delegates + reviews). Same architecture, different framings.