agent-orchestration - Agent Orchestration Patterns and Workflows
Patterns for building and coordinating AI agents including loops, multi-agent systems, framework selection, and scenario orchestration
Tags
Updated: 2026-03-24Capabilities
Typical Inputs
Typical Outputs
What this skill does
- implement ReAct pattern
- create reasoning loops
- build autonomous agents
- create execution plan
- execute task steps
- route tasks to agents
- run agents in parallel
- aggregate agent outputs
- resolve agent conflicts
- synthesize agent results
- build role-based teams
- implement hierarchical coordination
- use framework-specific workflows
- compare framework capabilities
- select appropriate framework
- run skill across scenarios
- execute parallel scenarios
- aggregate scenario results
Inputs
- agent questions
- task content
- tool definitions
- agent definitions
- task specifications
- framework requirements
- workload data
Outputs
- text response
- execution plan
- agent observations
- task results
- agent reports
- synthesized summary
- arbitration outcome
- framework selection
- scenario results
Requirements
- Claude Code 2.1.76+
- LLM API access
- Async execution environment