skills.rest
hlibkoval
hlibkovalCOMMUNITY
Explore all skills by this author →
GitHub

sub-agents-doc

Guide creating and configuring Claude Code subagents with frontmatter and tool access.

1|Updated Feb 17, 2026
Search Tags:#documentation#permissions#subagents#memory#claude-code#tools#frontmatter
One-click install
npx skills add https://github.com/hlibkoval/claudemd --skill sub-agents-doc
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: sub-agents-doc
Source: https://github.com/hlibkoval/claudemd/tree/main/skills/sub-agents-doc
Command: npx skills add https://github.com/hlibkoval/claudemd --skill sub-agents-doc

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Provides official guidance for creating, configuring, and operating Claude Code subagents, including frontmatter options, memory scopes, tool access, hooks, and fork modes.

Core Features & Use Cases

  • Official coverage of built-in and custom subagents, frontmatter fields, and lifecycle hooks.
  • Guidance on tool access, memory scopes (user, project, local), MCP servers, and permission modes.
  • Patterns for deploying subagents across project and user scopes, and forking workflows.

Quick Start

Create a new subagent file with YAML frontmatter in the appropriate agents folder and provide a concise system prompt in the body.

Frequently Asked Questions about sub-agents-doc

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

FAQPage Schema
How do I create a custom Claude Code subagent?▼

To create a custom Claude Code subagent, add a new file with YAML frontmatter to the appropriate agents folder and provide a concise system prompt in the body to define its behavior and configuration.

What frontmatter fields are available for configuring Claude Code subagents?▼

Claude Code subagents support frontmatter fields for managing tool access, permission modes, persistent memory, hooks, and MCP server configurations across project, user, and plugin scopes.

How do subagent memory scopes work in Claude Code?▼

Subagent memory scopes operate across user, project, and local levels, allowing persistent context retention. You can configure these scopes within the subagent frontmatter to manage data isolation and sharing.

Can I preload skills and manage MCP servers for Claude Code subagents?▼

Yes, you can preload skills and manage MCP servers for subagents. The official documentation describes how to configure these settings to operate subagents within different environments and scopes.

What is fork mode and how does it apply to Claude Code subagents?▼

Fork mode is a deployment pattern for Claude Code subagents that enables specialized forking workflows. It operates alongside permission modes and lifecycle hooks to manage subagent execution across different settings.

How do I restrict tool access and permissions for a Claude Code subagent?▼

You restrict tool access and permissions for a Claude Code subagent by defining specific permission modes within the YAML frontmatter, ensuring the subagent only utilizes explicitly authorized tools and operations.

Related Skills

Explore More Software Engineering Skills→
mdc159mdc159

opencode-folder-creator

Generate .opencode folder structures with opencode.json config and agent directories.

Community
Basic
iVintikiVintik

diagnose

Diagnose Claude Code plugin skill load, trigger, and runtime failures.

Community
Intermediate
LamseyahEliasLamseyahElias

hermes-agent

Configure and extend Hermes Agent for multi-agent orchestration across gateways.

Community
Advanced
xxihxxih
1

subagent-driven-development

Dispatch fresh subagents for isolated task execution with two-stage reviews.

Community
Advanced