agent-creator

Create Claude Code subagents as Markdown files with YAML frontmatter.

Updated Dec 18, 2025
One-click install
npx skills add https://github.com/alexjx/skills --skill agent-creator-alexjx
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: agent-creator
Source: https://github.com/alexjx/skills/tree/main/.claude/skills/agent-creator
Command: npx skills add https://github.com/alexjx/skills --skill agent-creator-alexjx

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps you create Claude Code subagents — modular, specialized AI assistants defined as Markdown files with YAML frontmatter. It standardizes how agents are described and invoked, enabling consistent delegation across tasks.

Core Features & Use Cases

  • Subagent format: Define a specialized AI assistant in a Markdown file with YAML frontmatter.
  • Trigger-driven delegation: Claude Code routes tasks to the most appropriate subagent based on its description.
  • Extensible customization: Specify tools, model, and output style per subagent, making it easy to create focused copilots.

Quick Start

Create a new subagent by adding a Markdown file under .claude/agents/ named after your agent and include a concise description, tool requirements, and a system prompt in YAML frontmatter.

Frequently Asked Questions about agent-creator

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

FAQPage Schema
How do I create Claude Code subagents with YAML frontmatter?

Create a Markdown file under .claude/agents/ with YAML frontmatter containing name and description fields, followed by a system prompt. Store the file in .claude/agents/ or ~/.claude/agents/ to enable Claude Code to recognize and route tasks to your subagent based on its description.

What's the difference between subagents and regular Claude Code workflows?

Subagents are specialized, modular AI assistants defined as Markdown files with standardized YAML frontmatter that enable trigger-driven delegation. They route tasks automatically based on description keywords, whereas regular workflows require manual invocation and lack the structured format for consistent specialization.

Can I specify custom tools and models for individual subagents?

Yes, subagents support optional tools and model fields in YAML frontmatter. This lets you customize which tools each subagent can access and which model it uses, tailoring specialized assistants for code review, debugging, documentation, or automation tasks.

How do trigger words activate the right subagent?

Include trigger words in your subagent's description field within the YAML frontmatter. Claude Code routes incoming tasks to the subagent whose description best matches the request, enabling context-aware delegation without manual routing logic.

What file structure do subagents require?

Subagents are Markdown files stored under .claude/agents/ or ~/.claude/agents/ with YAML frontmatter at the top containing name and description (required), tools and model (optional), followed by a focused system prompt in the Markdown body.

Can subagents handle automation tasks within Claude Code?

Yes, subagents excel at automation tasks. Design them with specific tools and prompts for code review, debugging, documentation, or custom automation workflows, then trigger them through Claude Code by matching task descriptions to subagent trigger words.