agent-creator

Create and validate Claude Code custom agent markdown files with frontmatter and tool configuration.

Updated Mar 22, 2026
One-click install
npx skills add https://github.com/kubrickcode/lingo-node --skill agent-creator-kubrickcode
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: agent-creator
Source: https://github.com/kubrickcode/lingo-node/tree/main/.agents/skills/agent-creator
Command: npx skills add https://github.com/kubrickcode/lingo-node --skill agent-creator-kubrickcode

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill helps you create well-scoped, effective custom agents for Claude Code without guessing at frontmatter fields, prompt structure, tool permissions, or model choices.

Core Features & Use Cases

  • Agent Definition Guidance: Explains how to write agent frontmatter with strong descriptions, explicit tool scoping, model selection, and optional advanced fields.
  • Prompt Structure Templates: Provides clear patterns for minimal, standard, and comprehensive agent system prompts based on domain complexity.
  • Initialization and Validation: Includes scripts to generate a new agent template and validate naming, frontmatter, trigger language, tool choices, and body structure.
  • Use Case: When you need a specialized subagent such as a database optimizer, documentation writer, or research analyst, this Skill helps you define the agent consistently and safely.

Quick Start

Ask the agent-creator skill to create or refine a Claude Code custom agent for your target domain and validate its frontmatter, tool scope, and system prompt structure.

Frequently Asked Questions about agent-creator

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

FAQPage Schema
How do I create a custom Claude Code subagent with correct frontmatter and tool permissions?

To create a custom Claude Code subagent, you need to define a markdown file with proper YAML frontmatter, a trigger-oriented description, an explicit tool allowlist, and a scoped system prompt. This ensures the agent operates safely within its designated domain.

What is the best way to structure a system prompt for a domain-specialist AI agent?

Structuring a domain-specialist AI agent system prompt depends on complexity, ranging from minimal to comprehensive templates. Effective prompt design requires clear instructions, explicit tool scoping, and kebab-case naming to ensure the agent triggers correctly for specialized tasks.

Does Claude Code agent creation support validating YAML frontmatter and tool allowlists?

Yes, Claude Code agent creation includes initialization and validation scripts that check YAML frontmatter, kebab-case naming, trigger language, and explicit tool allowlists. This validation ensures custom agents meet required model and permission specifications before deployment.

Can I update existing Claude Code agent specifications without breaking their tool scoping?

You can update existing Claude Code agent specifications by refining the system prompt structure and tool configuration. The validation workflow checks model choices and permissions, ensuring updates maintain safe and correctly scoped tool access for the specialized subagent.

When do I need to use kebab-case naming and explicit tool allowlists for custom agents?

Kebab-case naming and explicit tool allowlists are required whenever you create or refine a custom agent in Claude Code. These constraints ensure consistent agent identification and prevent unauthorized tool access during domain-specific task execution.