skills.rest
vietbui1999ru
vietbui1999ruCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

agent-orchestration

Design multi-agent workflows with roles, routing, and harness rules.

2|Updated Apr 21, 2026
Search Tags:#multi-agent#workflow-design#decision-tree#harness-design#agent-orchestration#team-coordination#automation-patterns
One-click install
npx skills add https://github.com/vietbui1999ru/llm-wiki --skill agent-orchestration-vietbui1999ru
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: agent-orchestration
Source: https://github.com/vietbui1999ru/llm-wiki/tree/main/claude-setup/skills/agent-orchestration
Command: npx skills add https://github.com/vietbui1999ru/llm-wiki --skill agent-orchestration-vietbui1999ru

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Coordinating complex multi-agent systems often requires clear patterns, roles, and harnesses. This skill consolidates orchestration patterns, decision criteria, and best practices for designing agent workflows and routing delegation.

Core Features & Use Cases

  • Guidance on when to use agents, subagents, or teams to partition work and coordinate results
  • Rules for subagent design, memory, isolation, and explicit skills for domain knowledge
  • Harness principles to ensure deterministic outcomes, verifiability, and safe recovery

Quick Start

Outline an orchestration plan where three subagents partition a task and report results to the lead agent.

Frequently Asked Questions about agent-orchestration

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

FAQPage Schema
How do I coordinate multiple AI agents to partition a complex task?▼

Multi-agent orchestration partitions work by assigning specific roles to subagents or teams, then routing their individual results back to a lead agent for deterministic synthesis and verification.

What is the best way to design a multi-agent workflow for task distribution?▼

Designing a multi-agent workflow requires defining agent roles, routing decisions, memory isolation, and harness verification rules to ensure deterministic outcomes and safe recovery during complex task execution.

When should I use subagents versus full agent teams for breaking up work?▼

Use subagents for isolated, domain-specific tasks requiring explicit skills and memory isolation, while deploying full agent teams when multiple roles need continuous coordination and complex routing decisions.

How do I build a harness system for safe multi-agent orchestration?▼

Build an orchestration harness by implementing verification principles that ensure deterministic outcomes, validate agent outputs, and enforce safe recovery protocols when subagents fail or produce unexpected results.

Can I enforce memory isolation across different agents in a coordinated workflow?▼

Memory isolation is enforced by applying specific orchestration rules that separate domain knowledge and task context, preventing subagents from interfering with each other during multi-agent workflow execution.

Why does my multi-agent workflow produce non-deterministic results?▼

Non-deterministic multi-agent workflows occur when harness principles lack strict verification rules, proper routing decisions, and memory isolation, preventing safe recovery and consistent task distribution outcomes.

Related Skills

Explore More Software Engineering Skills→
opensage-agentopensage-agent
113

workflow

Orchestrate multi-agent systems with ensemble, pool, and retry patterns.

Official
Advanced
mkalhitti-cloudmkalhitti-cloud
2

antigravity-core

Coordinate multiple AI perspectives into structured decision trees and synthesized outputs.

Community
Basic
Maheen-ZubairMaheen-Zubair

skill-creator

Create specialized Claude skills with structured development and packaging guidance.

Community
Advanced
vadimcomanescuvadimcomanescu
24

agents-crewai

Orchestrate role-based AI agents in sequential or hierarchical multi-step workflows.

Community
Advanced