write-a-skill

Create agent skill directories with SKILL.md files and YAML frontmatter.

1|1|Updated Apr 24, 2026
One-click install
npx skills add https://github.com/luan/agents --skill write-a-skill-luan
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: write-a-skill
Source: https://github.com/luan/agents/tree/main/skills/write-a-skill
Command: npx skills add https://github.com/luan/agents --skill write-a-skill-luan

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill helps users create new agent skills with consistent structure, clear progressive disclosure, and bundled resources so the skill is easy to discover and activate.

Core Features & Use Cases

  • Requirements gathering: Prompts for the skill’s domain, use cases, whether scripts are needed, and any reference materials to include.
  • Skill drafting: Produces a root SKILL.md with concise instructions and optionally adds supporting reference and example files when content grows.
  • Resource bundling: Recommends adding deterministic utility scripts for repeatable operations while avoiding tests that assert instruction wording.
  • Review alignment: Guides a user through validating coverage against intended use cases and refining detail level.

Quick Start

Ask to create a new skill by telling the domain, target user tasks, whether the skill needs scripts or only instructions, and any reference materials to bundle.

Frequently Asked Questions about write-a-skill

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

FAQPage Schema
How do I create agent instructions with YAML frontmatter?▼

To create agent instructions, you generate a root SKILL.md file with valid YAML frontmatter containing at least the name and description fields, followed by concise operator-authored guidance for the specific domain.

What is progressive disclosure in skill authoring?▼

Progressive disclosure in skill authoring is a structuring technique where you keep the root SKILL.md concise and move detailed domain references or examples into optional supporting files to ensure easy discovery and activation.

When should I include deterministic scripts in my agent workflow design?▼

You should include deterministic scripts in your agent workflow design for repeatable operations that require consistent execution, ensuring the skill remains reliable without relying solely on natural language instruction interpretation.

Do I need to write contract-style tests for skill instructions?▼

No, you do not need to write contract-style tests that assert instruction wording when authoring skills. Safe structuring avoids these tests and focuses on clear, concise operator-authored guidance instead.

What is the best way to structure a new agent skill directory?▼

The best way to structure an agent skill directory is to start with a root SKILL.md for core instructions, then add optional supporting reference and example files as content grows to maintain clear progressive disclosure.