Can I define custom handoff contracts between agents?

Yes, the skill is designed to help you define explicit handoff contract fields for every edge in your workflow to ensure data integrity between agents.

What workflow patterns does this skill support?

It supports five core patterns: Sequential (step-by-step), Parallel (fan-out/in), Router (intent-based dispatch), Orchestrator (central planner), and Evaluator (generator + quality gate).

When should I use a multi-agent workflow instead of a single prompt?

Use workflows when a single prompt cannot handle task complexity, when you need specialist agents with clear boundaries, or when you require deterministic validation loops.

How do I generate a new workflow skeleton?

You can use the provided workflow_scaffolder.py script via the command line to generate skeletons for specific patterns, such as 'sequential' or 'orchestrator'.

How does this skill manage AI costs?

It implements context and cost discipline by encouraging targeted artifacts instead of full context passing and by enforcing budget/timeout limits at every step.

Agent Workflow Designer

Name: Agent Workflow Designer
Author: zhangzhang-111-i

byzhangzhang-111-i

0•

Ciencia de Datos y ML

Orchestrates complex multi-agent systems using production-grade patterns, automated scaffolding, and robust error-handling strategies.

The Agent Workflow Designer skill enables developers to architect sophisticated AI systems by moving beyond single-prompt solutions to structured multi-agent workflows. It provides tools for generating skeleton configurations based on proven patterns like sequential chains, orchestrators, and parallel processing, while emphasizing context management, cost discipline, and explicit handoff contracts to ensure reliability and safety in production environments.

Características Principales

010 GitHub stars

02Pre-defined error recovery and retry strategy scaffolding

03Built-in context and cost discipline for long-running agent flows

04Multi-pattern selection including Sequential, Parallel, Router, and Orchestrator

05Detailed pattern maps for selecting the right dependency structure

06Automated skeleton configuration generation for rapid bootstrapping

Casos de Uso

01Implementing safety and quality gates using generator-evaluator loops

02Building complex content pipelines requiring sequential specialist reviews

03Developing intelligent incident triage systems using the orchestrator pattern

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add zhangzhang-111-i/claude-skills111 agent-workflow-designer

For use in Claude.ai and ChatGPT

Características Principales

010 GitHub stars

02Pre-defined error recovery and retry strategy scaffolding

03Built-in context and cost discipline for long-running agent flows

04Multi-pattern selection including Sequential, Parallel, Router, and Orchestrator

05Detailed pattern maps for selecting the right dependency structure

06Automated skeleton configuration generation for rapid bootstrapping

Casos de Uso

01Implementing safety and quality gates using generator-evaluator loops

02Building complex content pipelines requiring sequential specialist reviews

03Developing intelligent incident triage systems using the orchestrator pattern

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add zhangzhang-111-i/claude-skills111 agent-workflow-designer

For use in Claude.ai and ChatGPT