Multi-Agent System
Also known as: MAS, AI agent team, Agent crew
A system where multiple specialized AI agents collaborate on tasks.
What is Multi-Agent System?
A multi-agent system (MAS) is a configuration of two or more AI agents that work together — often each specialized for a sub-task — sharing memory, routing tasks by capability, and coordinating handoffs. Production multi-agent systems typically run on an agent operations fabric like LeafMesh ADK that handles inter-agent state, policy, and observability.
Related concepts
AI Agent TeamAgent Orchestration
Looking for the agent operations fabric?
LeafMesh ADK is the production runtime for the concepts defined here.
More definitions
Agent Operations FabricAI Agent OrchestrationAutonomous AIAgentic AIAI Control PlaneAgent Development Kit (ADK)Human-in-the-Loop (HITL)AI GovernanceShared Memory (Agent Context)YAML-First ConfigurationEscalation RoutingAgent ObservabilityVendor-Agnostic AILLM OrchestrationAgentOpsAgent Operations Platform