skill-creator

Plan and implement Claude Skills with progressive disclosure and deterministic evaluation scripts.

Updated Apr 19, 2026
One-click install
npx skills add https://github.com/R3n3gade-ai/canvasclaw --skill skill-creator-r3n3gade-ai
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: skill-creator
Source: https://github.com/R3n3gade-ai/canvasclaw/tree/main/jiuwenclaw/resources/agent/jiuwenclaw_workspace/skills/skill-creator
Command: npx skills add https://github.com/R3n3gade-ai/canvasclaw --skill skill-creator-r3n3gade-ai

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires yaml, and includes scripts (resource) components.

What problem does it solve?

This skill helps teams design, test, and iteratively improve Claude Skills by providing a clear framework for drafting SKILL.md, running deterministic evaluations, and recording results for auditing and reuse.

Core Features & Use Cases

  • Structured skill planning guides users from intent capture to frontmatter and body instructions.
  • Deterministic tooling bundles scripts for evaluation, benchmarking, and description improvement.
  • Domain organization supports multi‑domain variants with references and assets as needed.
  • Evidence-based iteration keeps a history of results to drive automated improvements.

Quick Start

Immediately craft a first SKILL.md draft and run the evaluation loop to start iterating toward a trigger-ready skill.

Frequently Asked Questions about skill-creator

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

FAQPage Schema
How do I create production-ready Claude Skills from scratch?

Claude Skills require a structured SKILL.md file with frontmatter and body instructions. This skill guides you through planning, drafting, testing, and iteratively improving skills to ensure they are production-ready and trigger accurately.

What is progressive disclosure in skill design?

Progressive disclosure in skill design maps keyword-based discovery to SKILL.md frontmatter and optional resources. It organizes modular structures and references to maximize trigger accuracy while maintaining safety and versioning throughout the skill lifecycle.

How do I evaluate and benchmark Claude Skills effectively?

You evaluate Claude Skills by running bundled deterministic scripts for testing and benchmarking. The framework records evaluation results to maintain an evidence-based history, driving automated improvements and refining trigger accuracy over time.

Does skill creation require specific dependencies or formats?

Skill creation requires the yaml dependency for frontmatter formatting. You structure SKILL.md files with keywords, taglines, and modular components, mapping domain organization to frontmatter to ensure accurate keyword-based discovery.

What's the best way to iterate on SKILL.md trigger accuracy?

The best way to iterate on SKILL.md trigger accuracy is through an evaluation loop. You draft initial instructions, run deterministic benchmarking scripts, audit the results history, and apply evidence-based improvements to frontmatter and descriptions.

When should I use bundled scripts for skill automation?

Use bundled scripts for skill automation when you need deterministic evaluation, benchmarking, and description improvement. This orchestrates automated testing loops and maintains versioning safety for modular structures and optional resource references.