skill-creator

Create and package modular AgentSkills with SKILL.md frontmatter metadata.

58|1|Updated May 13, 2026
One-click install
npx skills add https://github.com/Simplified-Reasoning/Pi-Bench --skill skill-creator-simplified-reasoning
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: skill-creator
Source: https://github.com/Simplified-Reasoning/Pi-Bench/tree/main/third_party/nanobot/nanobot/skills/skill-creator
Command: npx skills add https://github.com/Simplified-Reasoning/Pi-Bench --skill skill-creator-simplified-reasoning

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps you design, structure, and package self-contained AgentSkills with clear triggering metadata and efficient context loading.

Core Features & Use Cases

  • Skill anatomy and requirements: Define the required SKILL.md frontmatter (name and description) and provide operational instructions in the body for when the skill triggers.
  • Progressive disclosure guidance: Keep SKILL.md concise and load deeper guidance via optional bundled resources (scripts, references, assets) only when needed.
  • Packaging and iteration workflow: Follow a practical process to initialize, implement, validate, package, and iterate on skills.

Quick Start

Use the skill to package an existing AgentSkill folder by running the packaging script against the skill directory.

Frequently Asked Questions about skill-creator

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

FAQPage Schema
How do I structure a modular agent skill for automation workflows?

To package an agent skill, run the packaging script against your skill directory to generate a distributable .skill package. You must ensure correct resource organization and validate naming rules before producing the final distributable asset.

What is progressive disclosure in agent framework skill design?

Progressive disclosure in skill design involves keeping the SKILL.md body concise and loading deeper guidance only when needed via optional bundled resources like scripts, references, and assets. This optimizes context loading across diverse agent workflows.

Do I need YAML frontmatter to build a reusable agent skill?

Yes, YAML frontmatter is required to build a reusable agent skill. You must define the required name and description metadata in the SKILL.md frontmatter to establish clear skill triggers and ensure proper validation before packaging.

Can I bundle deterministic scripts and reference documents inside an agent skill?

Yes, you can bundle deterministic scripts, reference documents, and reusable assets inside an agent skill. These optional bundled resources support progressive disclosure by loading deeper guidance into the agent workflow only when needed.

What is the best way to validate agent skill metadata before distribution?

The best way to validate agent skill metadata is to follow the packaging and iteration workflow, checking adherence to naming and validation rules against the SKILL.md frontmatter. This ensures required metadata is correct before producing a distributable .skill package.