---
license: mit
language:
- en
- zh
- ja
- ko
pretty_name: "Agent Workflow Playbook"
tags:
- agent-design
- agent-framework
- agentic-ai
- ai-agent
- automation
- clawhub
- developer-tools
- distribution
- marketplace
- multi-agent
- orchestration
- plugin
- plugin-marketing
- skill-architecture
- workflow
size_categories:
- n<1K
---

# Agent Workflow Playbook

> AI Agent Workflow & Skill Architecture Guide — Design patterns for building autonomous AI agent systems, plugin marketplaces, and multi-agent orchestration....

## 📦 Install on ClawHub

```bash
clawhub install agent-workflow-playbook
```

Then ask your AI agent:

> "Design an autonomous AI agent workflow for our customer support"

Installs the full **Agent Workflow Playbook** playbook — battle-tested with 30+ Product Hunt #1 wins, AFFiNE 60K+ GitHub stars, and 150+ AI startup consultations.

🔗 [View on ClawHub](https://clawhub.ai/gingiris/agent-workflow-playbook) · [All Gingiris skills](https://clawhub.ai/user/gingiris) · [Author blog](https://gingiris.tools)

## 🚀 30-Second Preview

```
You: Design an autonomous AI agent workflow for our customer support
Agent (with agent-workflow-playbook):
  Stage 1: intent classification (ticket → category, confidence threshold ≥0.85)
  Stage 2: knowledge retrieval (semantic search over docs + past tickets)
  Stage 3: response draft (humans review for high-stakes categories)
  Stage 4: feedback loop (every escalation re-trains the classifier weekly)
```

Concrete framework, real case studies, decision rules — not generic advice.

---

## 📚 Full Playbook

---
name: agent-workflow-playbook
description: |
  AI Agent Workflow & Skill Architecture Guide — Design patterns for building autonomous AI agent systems, plugin marketplaces, and multi-agent orchestration. Battle-tested framework from shipping 200+ skills on ClawHub. By @WeiYipei.
---

# AI Agent Workflow Playbook — Skill Architecture & Plugin Marketing

> 🌍 **Language / 语言**: [中文](#中文版) | [English](references/en/README.md) | [日本語](references/ja/README.md) | [한국어](references/ko/README.md)

## 📦 Install

```bash
clawhub install agent-workflow-playbook
```

**What you get after installing:**
- End-to-end AI agent workflow design patterns
- Multi-agent orchestration frameworks (sequential, parallel, hierarchical)
- Production deployment checklist for agent systems

---

Complete guide for designing AI agent automation systems. Covers skill/plugin architecture, multi-agent coordination patterns, marketplace distribution strategies, and workflow orchestration best practices.

## What You'll Learn
- Skill architecture patterns (single-purpose vs. orchestrator agents)
- Plugin marketplace SEO and distribution
- Multi-agent workflow design (sequential, parallel, hierarchical)
- Agent-to-agent communication protocols
- Monetization strategies for AI agent skills

## Who This Is For
- AI developers building agent systems
- Teams publishing skills/plugins to marketplaces
- Product managers designing agent-native products

Built by Gingiris — shipped 200+ skills, 10k+ installs across ClawHub.


---

*Part of the [Gingiris growth skills suite](https://clawhub.ai/user/gingiris) — battle-tested operational playbooks for AI startups. Author: Iris (生姜iris), ex-AFFiNE COO, Forbes Asia 30 Under 30.*