Category:Multi-Agent System
Jump to navigation
Jump to search
This report lists the pages about multi-agent systems in GM-RKB.
A Multi-Agent System is an arrangement in which several agents work together on a task that no single agent completes alone.
Member Kinds
The members fall into four kinds.
- Orchestration frameworks: AutoGen Framework, CrewAI Framework, MetaGPT Framework, Multi-Agent Development Framework, AI Agent Orchestration Framework.
- Coordination patterns: Multi-Agent Orchestration, Parallel AI Agent Orchestration Pattern, AI Agent Parallelization System, Consensus-Based Evaluation Mechanism.
- Sub-agent structures: Agentic Coding Subagent, Deep Agent Sub-Agent and Subagent System, where one agent delegates to others it creates.
- Governed collectives: Agent Fleet, Fleet (AGET) and GM-RKB Fleet, where the agents are long-lived seats under a shared governance policy rather than tasks spawned on demand.
The distinction between the last two kinds is the useful one: a sub-agent structure is transient and task-shaped, while a governed collective persists and is administered.
Protocols by which members communicate are here too, including Agent-to-Agent Collaboration Protocol and Multi-Agent System Protocol.
See Also
Pages in category "Multi-Agent System"
The following 27 pages are in this category, out of 27 total.
A
M
- MetaGPT Framework
- Multi-Agent AI System
- Multi-Agent AI Workflow Framework
- Multi-Agent Development Framework
- Multi-Agent Orchestration
- Multi-Agent Orchestration Framework
- Multi-Agent Physics Training Environment
- Multi-Agent Research Pipeline
- Multi-Agent Skill Orchestration System
- Multi-Agent System Architecture
- Multi-Agent System Protocol
- Multi-Change Redlining Agent System