write-a-skill

Create agent skills with SKILL.md entry points and supporting resources.

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

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill provides a repeatable process to convert user requests or demonstrated workflows into self-contained agent skills with a proper SKILL.md entry point, ensuring consistent discovery and predictable behavior.

Core Features & Use Cases

  • Requirements Gathering: Prompts to capture task scope, use cases, and whether deterministic scripts or references are needed.
  • Skill Drafting & Structure: A template for SKILL.md, guidance on when to add scripts/, references/, or assets/, and recommendations for splitting content.
  • Review & Validation: A checklist to ensure the description includes explicit triggers, avoids time-sensitive info, and keeps references one level deep.
  • Use Case: Convert a demonstrated data-cleaning workflow into a reusable skill that includes SKILL.md, example workflows, and a scripts/ folder for repeatable transforms.

Quick Start

Create a new SKILL.md named "analytics-report" with a frontmatter name and description that includes trigger phrases, add a short Quick Start workflow and a scripts/ directory for any deterministic transforms.

Frequently Asked Questions about write-a-skill

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

FAQPage Schema
How do I convert a workflow into a reusable agent skill?

To convert a workflow into a reusable agent skill, you package the demonstrated steps into a self-contained SKILL.md entry point with required YAML frontmatter, optional deterministic scripts, and reference files for predictable agent behavior.

What is the required structure for a SKILL.md entry point?

A SKILL.md entry point requires YAML frontmatter containing name and description fields, with the description featuring explicit trigger wording to ensure proper agent discovery and consistent execution.

When should I add scripts or references directories to an agent skill?

You should add scripts/, references/, and assets/ directories when packaging agent skills that require deterministic code execution, supporting documentation, or static files to reliably complete the targeted workflow.

Can I turn a user prompt directly into a self-contained agent capability?

Yes, you can convert user prompts into reusable capabilities by drafting a structured SKILL.md file that captures the task scope, trigger phrases, and any necessary deterministic scripts for the agent to execute.

What validation checklist should I follow when authoring agent skills?

When authoring agent skills, validate that the description includes explicit triggers, avoids time-sensitive information, splits content appropriately, and keeps references structured at most one level deep.