Agent Development

Create Claude Code agent files with YAML frontmatter and structured system prompts.

1|Updated May 4, 2026
One-click install
npx skills add https://github.com/open-session/os_design_system --skill agent-development-open-session
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: Agent Development
Source: https://github.com/open-session/os_design_system/tree/main/.claude/plugins/plugin-dev/skills/agent-development
Command: npx skills add https://github.com/open-session/os_design_system --skill agent-development-open-session

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

This Skill helps you design and package autonomous Claude Code agents with the correct frontmatter, trigger logic, and system prompt structure so they are easy to discover, validate, and use.

Core Features & Use Cases

  • Agent Frontmatter Design: Create valid agent metadata with the required name, description, model, color, and optional tools fields.
  • Trigger Engineering: Write strong description text with concrete <example> blocks so the right agent activates in the right scenarios.
  • System Prompt Architecture: Shape the agent’s body into a clear operating manual with responsibilities, process steps, quality standards, and output format.
  • Validation and Testing: Use the bundled validation script and reference guidance to check structure, improve reliability, and catch formatting mistakes before deployment.
  • Use Case: A plugin developer wants a code-review agent that triggers after code is written; this Skill provides the template, guidance, and checks needed to produce a dependable agent file.

Quick Start

Use the agent development skill to create a Claude Code agent for your task with clear triggers, a structured system prompt, and valid frontmatter.

Frequently Asked Questions about Agent Development

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

FAQPage Schema
How do I write a system prompt for a Claude Code agent?

To write a system prompt for a Claude Code agent, structure the body as an operating manual with responsibilities, process steps, quality standards, and output formats. This structured approach ensures the autonomous agent operates reliably.

What is the best way to format frontmatter for Claude Code agents?

Formatting frontmatter for Claude Code agents requires YAML syntax with name, description, model, color, and optional tools fields. Correct frontmatter design ensures the agent file is valid, discoverable, and ready for validation.

How do I add trigger examples to a plugin dev agent?

Add trigger examples to a plugin dev agent by writing strong description text with concrete example blocks. This trigger engineering ensures the right agent activates automatically in the correct scenarios.

Can I validate a Claude Code agent file before deployment?

Yes, you can validate a Claude Code agent file using a bundled validation script and reference guidance. This checks structure, improves reliability, and catches formatting mistakes in the frontmatter and system prompt before deployment.

Why does my Claude Code agent not trigger automatically?

A Claude Code agent may fail to trigger automatically if the description lacks concrete example blocks. Engineering strong trigger logic with specific examples ensures the correct agent activates in the right scenarios.

Do I need YAML frontmatter to create a Claude Code agent?

Yes, YAML frontmatter is required to create a Claude Code agent. It provides the necessary discovery-ready metadata including name, description, model, and tool configuration for the agent to function properly.