subagent-creator

Create Claude Code sub-agents with custom system prompts and tool configurations.

Updated Aug 27, 2026
One-click install
npx skills add https://github.com/jeehoon0310/claude-code-config --skill subagent-creator-jeehoon0310
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: subagent-creator
Source: https://github.com/jeehoon0310/claude-code-config/tree/main/project/skills/subagent-creator
Command: npx skills add https://github.com/jeehoon0310/claude-code-config --skill subagent-creator-jeehoon0310

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

Create specialized Claude Code sub-agents with custom system prompts and tool configurations. Use when users ask to create a new sub-agent, custom agent, specialized assistant, or want to configure task-specific AI workflows for Claude Code.

Core Features & Use Cases

  • Define sub-agent metadata: name, description, tools, model, permissionMode, and optional skills
  • Write a structured system prompt that governs role, responsibilities, and behavior
  • Choose deployment scope: project-level (.claude/agents/) or user-level (e.g., ~/.claude/agents/)
  • Create and store the sub-agent as a Markdown file with a frontmatter header
  • Reuse templates and references to streamline creation and validation

Quick Start

Create a new sub-agent by saving a Markdown file under the chosen agents directory with the required frontmatter and an accompanying system prompt.

Frequently Asked Questions about subagent-creator

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

FAQPage Schema
How do I create a custom Claude Code sub-agent for automated code review?

Create a custom Claude Code sub-agent by saving a Markdown file with a structured YAML frontmatter header defining name, description, tools, and model, followed by a system prompt body governing its behavior. This enables specialized automated workflows like code review.

What is the YAML frontmatter schema required for defining Claude Code sub-agents?

The YAML frontmatter schema for Claude Code sub-agents requires name and description fields, with optional tools, model, permissionMode, and skills fields. This structured metadata enables auto-loading of templates and references for the agent.

Can I deploy sub-agents at both project level and user level in Claude Code?

Yes, you can deploy Claude Code sub-agents at project-level by saving files under the .claude/agents/ directory, or at user-level by saving them in the ~/.claude/agents/ directory. This allows flexible scoping for different automation needs.

How do I configure tool permissions and models for specialized Claude Code assistants?

Configure tool permissions and models for Claude Code assistants by setting the optional tools, model, and permissionMode fields in the sub-agent's YAML frontmatter. This restricts and defines the operational scope for tasks like debugging or data analysis.

Does Claude Code support reusing templates and references when creating sub-agents?

Yes, the sub-agent creation process supports reusing templates and references to streamline creation and validation. This ensures structured system prompts and consistent configurations across multiple specialized agents.

What are the limitations when defining system prompts for custom Claude Code sub-agents?

System prompts for custom Claude Code sub-agents must be written in a structured Markdown body following the YAML frontmatter. While the prompt content is flexible, the preceding metadata schema strictly enforces required fields to enable proper auto-loading.