npx claudepluginhub nwave-ai/nwave --plugin nwThis skill uses the workspace's default tool permissions.
```
Provides agent orchestration patterns for ReAct loops, multi-agent coordination, CrewAI/AutoGen/Swarm evaluation, and multi-scenario workflows. Use for autonomous agents or complex AI systems.
AI agent workflow patterns including ReAct agents, multi-agent systems, loop control, tool orchestration, and autonomous agent architectures. Use when building AI agents, implementing workflows, creating autonomous systems, or when user mentions agents, workflows, ReAct, multi-step reasoning, loop control, agent orchestration, or autonomous AI.
Provides Claude Code templates for agentic design patterns like prompt chaining, routing, reflection, tool use, planning, and multi-agent workflows using 4-layer stack. For LLM task decomposition.
Share bugs, ideas, or general feedback.
Is the agent doing a single focused task?
YES -> Does it need self-evaluation?
YES -> Reflection
NO -> ReAct (default for most agents)
NO -> Is it coordinating multiple agents?
YES -> Are tasks independent?
YES -> Parallel Orchestration
NO -> Are tasks sequential with dependencies?
YES -> Sequential Orchestration
NO -> Hierarchical (supervisor + workers)
NO -> Is it routing to one of several specialists?
YES -> Router
NO -> Does it need structured task decomposition?
YES -> Planning
NO -> ReAct (default)
General-purpose agents needing tool calling and iterative problem-solving.
Loop: Reason -> Select/execute action -> Observe result -> Repeat until done. When: Default pattern. Most specialist agents. Examples: software-crafter, researcher, troubleshooter.
Agent must evaluate and iteratively improve its own output.
Loop: Generate -> Review against criteria -> Identify gaps -> Refine -> Validate threshold met. When: Quality-critical outputs where first-draft insufficient (code review, architecture review, agent validation). Examples: agent-builder-reviewer, solution-architect-reviewer, software-crafter-reviewer.
Request classified and delegated to exactly one specialist.
Loop: Analyze request -> Classify -> Select specialist -> Delegate. When: Task dispatching, single path execution. Low overhead, fast routing. Examples: workflow-dispatcher, task-router.
Complex tasks requiring structured decomposition before execution.
Loop: Decompose into sub-tasks -> Sequence -> Allocate resources -> Execute with checkpoints. When: Multi-step implementations, migrations, large refactoring. Examples: project-planner, migration-coordinator.
Linear workflows with clear dependencies between stages.
Structure: Agent1 -> Output1 -> Agent2 -> Output2 -> Agent3 -> Result When: Pipeline workflows where each stage transforms previous output. Example: nWave waves: DISCUSS -> DESIGN -> DEVOPS -> DISTILL -> DELIVER.
Multiple independent analyses needed simultaneously.
Structure: Supervisor -> [Worker1, Worker2, Worker3] (concurrent) -> Aggregate results. When: Independent analyses, multi-aspect reviews, parallel risk assessment. Example: Multi-reviewer code review, parallel security + performance + correctness analysis.
Supervisor coordinates multiple worker agents dynamically.
Structure: Supervisor manages workers, routing tasks and aggregating results. When: Complex coordination where routing depends on intermediate results. Example: feature-coordinator supervising frontend/backend/database/testing specialists.
| Agent Role | Pattern | Rationale |
|---|---|---|
| Specialist (single domain) | ReAct | Tool-using, iterative task completion |
| Reviewer (-reviewer suffix) | Reflection | Must self-evaluate and iterate on critique |
| Wave orchestrator | Sequential | Clear dependency chain between phases |
| Multi-agent coordinator | Hierarchical | Dynamic task routing to specialists |
| Task dispatcher | Router | Classification and single-path delegation |