nw-design-patterns

Select and apply one of seven agent design patterns for multi-agent workflows.

Updated Mar 18, 2024
One-click install
npx skills add https://github.com/v1bh0r/precise-ledger-pro --skill nw-design-patterns-v1bh0r
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: nw-design-patterns
Source: https://github.com/v1bh0r/precise-ledger-pro/tree/main/nWave/skills/nw-design-patterns
Command: npx skills add https://github.com/v1bh0r/precise-ledger-pro --skill nw-design-patterns-v1bh0r

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Helps teams identify and apply the appropriate multi-agent design pattern to structure intelligent workflows and reduce coordination complexity.

Core Features & Use Cases

  • Pattern Decision Tree guidance that maps task characteristics to one of seven agentic patterns: ReAct, Reflection, Router, Planning, Sequential Orchestration, Parallel Orchestration, and Hierarchical.
  • Clear guidance for when to use each pattern in single-task, multi-agent coordination, and pipeline scenarios.
  • Use cases include architecture planning for AI assistants, multi-agent problem solving, and orchestration across tools and services.

Quick Start

Apply the decision-tree to determine the correct agent pattern for a given task and implement the chosen pattern in your multi-agent workflow.

Frequently Asked Questions about nw-design-patterns

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
What is the best way to structure multi-agent workflows for complex tasks?

The best way to structure multi-agent workflows is to apply a decision tree that maps task characteristics to one of seven agentic patterns, including ReAct, Router, and Hierarchical orchestration, reducing coordination complexity.

How do I choose the right agent design pattern for my AI architecture?

To choose the right agent design pattern, evaluate task characteristics against a structured decision tree to identify whether a single-task pattern like Reflection or a multi-agent pattern like Sequential Orchestration is optimal.

When should I use parallel orchestration over hierarchical patterns in pipeline orchestration?

Use parallel orchestration when independent tasks require concurrent execution, whereas hierarchical patterns suit scenarios needing layered task delegation and centralized coordination across multiple agents.

Can I use this decision tree for single-agent problem solving?

Yes, the decision tree supports single-agent scenarios by identifying appropriate patterns like ReAct and Reflection for task routing and architecture planning without requiring multi-agent coordination.

How does a router pattern coordinate multiple agents in a workflow?

The router pattern coordinates workflows by dynamically directing tasks to specialized agents based on input characteristics, clarifying coordination roles and data flows across the orchestrated pipeline.

What are the limitations of using predefined patterns for task routing?

Predefined patterns limit task routing to seven specific architectural models, meaning highly customized coordination requirements may fall outside the supported ReAct, Reflection, Router, Planning, and Orchestration patterns.