skills-ref


Reference library for Agent Skills.

> Note: This library is intended for demonstration purposes only. It is not meant to be used in production.

Installation

``bash
pnpm install
`

After installation, skills-ref CLI will be available.

Usage

$3

`bash


Validate a skill

skills-ref validate path/to/skill

Read skill properties (outputs JSON)

skills-ref read-properties path/to/skill

Generate XML for agent prompts

skills-ref to-prompt path/to/skill-a path/to/skill-b
`

`bash
PS C:\Users\admin> npm install -g skills-ref

added 4 packages in 2s
PS C:\Users\admin> skills-ref --help
Usage: cli [options] [command]

Reference library for Agent Skills

Options:
-V, --version output the version number
-h, --help display help for command

Commands:
validate
read-properties
to-prompt
help [command] display help for command
PS C:\Users\admin> skills-ref validate F:\ai-project\skills\skills\algorithmic-art
Valid skill: F:\ai-project\skills\skills\algorithmic-art
PS C:\Users\admin>
PS C:\Users\admin> skills-ref read-properties F:\ai-project\skills\skills\algorithmic-art
{
"name": "algorithmic-art",
"description": "Creating algorithmic art using p5.js with seeded randomness and interactive parameter exploration. Use this when users request creating art using code, generative art, algorithmic art, flow fields, or particle systems. Create original algorithmic art rather than copying existing artists' work to avoid copyright violations.",
"license": "Complete terms in LICENSE.txt"
}
PS C:\Users\admin> skills-ref to-prompt F:\ai-project\skills\skills\algorithmic-art



algorithmic-art


Creating algorithmic art using p5.js with seeded randomness and interactive parameter exploration. Use this when users request creating art using code, generative art, algorithmic art, flow fields, or particle systems. Create original algorithmic art rather than copying existing artists' work to avoid copyright violations.


F:\ai-project\skills\skills\algorithmic-art\SKILL.md



PS C:\Users\admin>
`

$3

`typescript
import { validate, readProperties, toPrompt } from "skills-ref";

// Validate a skill directory
const problems = await validate("my-skill");
if (problems.length > 0) {
console.log("Validation errors:", problems);
}

// Read skill properties
const props = await readProperties("my-skill");
console.log(Skill: ${props.name} - ${props.description});

// Generate prompt for available skills
const prompt = await toPrompt(["skill-a", "skill-b"]);
console.log(prompt);
`

Development

`bash


Build the project

pnpm build

Run tests

pnpm test

Format code

pnpm format

Lint code

pnpm lint
pnpm lint:fix
`

Agent Prompt Integration

Use to-prompt to generate the suggested XML block for your agent's system prompt. This format is recommended for Anthropic's models, but Skill Clients may choose to format it differently based on the model being used.

`xml



my-skill


What this skill does and when to use it


/path/to/my-skill/SKILL.md



`

The ` element tells the agent where to find the full skill instructions.

License

MIT License