Multi-agent system
A multi-agent system is a set of agents that collaborate or hand off work, each with a role, tools and boundaries.
How it works
Split roles (research, draft, verify) only when a single agent becomes unmaintainable. Shared memory and clear contracts matter more than agent count. Start with one agent; split when eval proves the need.
Example
One agent retrieves policy docs, another drafts a reply, a third checks citations before human approval.