> ## Documentation Index > Fetch the complete documentation index at: https://learn.getodin.ai/llms.txt > Use this file to discover all available pages before exploring further. # Agent Communication Toolkit > Multi-Agent Systems - delegate tasks to specialized worker agents Enables Multi-Agent Systems (MAS). Allows a "Manager" agent to delegate tasks to specialized "Worker" agents. This toolkit transforms a single-agent system into a coordinated multi-agent architecture where different agents with specialized capabilities can work together to solve complex problems. Agent Communication configuration with target agent selection ## Key Features * **Agent Selection**: Intelligent routing mechanism that analyzes task requirements and selects the most appropriate specialized agent for the job. * **Context Passing**: Seamlessly transfers conversation history and relevant context between agents to maintain continuity. * **Response Integration**: Aggregates outputs from multiple worker agents into a coherent response for the user. * **Fallback Handling**: If a specialist agent fails, the system can route to alternative agents or escalate to human intervention. * **Agent Discovery**: Automatically detects available agents in the workspace and their capabilities. * **Task Orchestration**: Coordinates complex workflows involving multiple agents working in sequence or parallel. ## Design Patterns * **Hub and Spoke (Router)**: Central coordinator routes requests to specialized agents. * **Sequential Chain (Pipeline)**: Tasks flow through a sequence of agents. * **Parallel Processing (Fan-Out)**: Multiple agents work simultaneously on subtasks. * **Hierarchical (Tree)**: Multi-level delegation with supervisory agents. * **Expert Panel (Committee)**: Multiple specialists provide opinions; coordinator synthesizes.