agent-authoring

Design Claude Code subagents with consistent frontmatter and system prompts.

Updated May 12, 2026
One-click install
npx skills add https://github.com/RorySullivan1/claudeBrain --skill agent-authoring-rorysullivan1
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: agent-authoring
Source: https://github.com/RorySullivan1/claudeBrain/tree/main/.claude/skills/agent-authoring
Command: npx skills add https://github.com/RorySullivan1/claudeBrain --skill agent-authoring-rorysullivan1

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Crafting Claude Code subagents is error-prone without a consistent frontmatter schema and a clear mandate. This skill provides a structured template and guidance to design, write, scaffold, review, or fix subagents that live in .claude/agents/<name>.md.

Core Features & Use Cases

  • Frontmatter schema with required fields (name, description) and optional fields like tools, permissionMode, model.
  • A complete body that serves as the system prompt for the subagent, ensuring self-contained behavior.
  • An authoring checklist and template to standardize creation across projects and teams.

Quick Start

Create a new subagent file at .claude/agents/<name>.md using the provided YAML frontmatter template and fill in the mandate, tools, permissions, and model.

Frequently Asked Questions about agent-authoring

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

FAQPage Schema
How do I write a Claude Code subagent with proper frontmatter?

To write a Claude Code subagent, create a Markdown file at .claude/agents/<name>.md using a structured YAML frontmatter schema with required name and description fields, plus optional tools, permissionMode, and model configurations.

What fields are required in a Claude subagent frontmatter schema?

The required fields in a Claude subagent frontmatter schema are name and description, while optional fields include tools, permissionMode, and model to enforce tool usage constraints and model selection across projects.

How do I create a self-contained system prompt for a subagent?

A self-contained system prompt for a subagent is created by writing a complete body in the Markdown file that defines its mandate and behavior, ensuring the subagent operates independently without requiring external context.

What is the best way to standardize subagent creation across projects?

The best way to standardize subagent creation across projects is to use a reusable frontmatter schema combined with a guided template and an authoring checklist, ensuring consistent quality and clear mandates for every subagent.

Can I restrict tool usage and set model preferences for Claude subagents?

Yes, you can restrict tool usage and set model preferences for Claude subagents by defining the optional tools, permissionMode, and model fields within the YAML frontmatter schema of the subagent file.

Why does my Claude subagent behave inconsistently across different projects?

Inconsistent subagent behavior often results from lacking a clear mandate or a standardized frontmatter schema, which an authoring checklist and a self-contained system prompt template are designed to resolve.