Why does the skill require Context7 research during creation?

Context7 research is mandatory to ensure the agent is built using the most current Anthropic documentation, YAML schema requirements, and latest best practices.

How does this skill handle agent isolation and security?

It defines strict workspace boundaries and file access permissions to prevent scope creep, ensuring agents only operate within their designated folder structures.

Can I use the Agent Creator for existing agents?

Yes, it includes specific refactoring protocols to take overly verbose or legacy agents and restructure them into the modern, minimal skill-based format.

What are the mandatory sections of an agent created with this skill?

Every agent includes four core sections: Identity & Role, Iterative Workflow, Mandatory Skill Invocation, and a Quick Reference guide.

What is the recommended length for a Claude Code agent file?

This skill enforces a 'Minimal Agent' philosophy, aiming for approximately 150 lines by moving heavy technical procedures into separate companion SKILL.md files.

Claude Agent Creator

Name: Claude Agent Creator
Author: fercracix33

byfercracix33

0•

Herramientas para Desarrolladores

Architects and refactors Claude Code agents using Anthropic best practices for modular, minimal, and high-performance agent structures.

The Agent Creator skill streamlines the development of specialized Claude Code agents by enforcing a strict separation of concerns between agent identity and technical procedures. It guides developers through a structured 6-phase workflow—from initial discovery and mandatory Context7 documentation research to YAML frontmatter design and skill scaffolding. By following this skill's guidance, developers can ensure every agent remains lightweight (approximately 150 lines) while maintaining robust role boundaries, clear authority, and mandatory skill invocation patterns that align with official Anthropic recommendations.

Características Principales

01Automated skill directory scaffolding and metadata validation

02Automated agent refactoring workflows to maintain minimal core files

03Standardized YAML frontmatter generation with example-driven descriptions

04Integrated Context7 research checkpoints for latest Anthropic best practices

050 GitHub stars

06Strict role boundary definition and authority/prohibition mapping

Casos de Uso

01Building new specialized agents for specific domains like Security, UI/UX, or Databases

02Validating existing agent configurations against official naming and schema conventions

03Refactoring bloated, monolithic agent files into efficient skill-based modular structures

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add fercracix33/poli2-clean-architecture agent-creator

For use in Claude.ai and ChatGPT

Download Skill