skills.rest
loongclaw-ai
loongclaw-aiOFFICIAL
Explore all skills by this author →
Website𝕏@loongclawaiGitHub

skill-creator

Automate Claude skill creation with intent interviews, draft writing, and evaluation loops.

641|104|Updated Mar 5, 2026
Search Tags:#automation#benchmarking#claude#skill-creation#skill-evaluation#iterative-improvement#description-optimization
One-click install
npx skills add https://github.com/loongclaw-ai/loongclaw --skill skill-creator-loongclaw-ai
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: skill-creator
Source: https://github.com/loongclaw-ai/loongclaw/tree/main/skills/skill-creator
Command: npx skills add https://github.com/loongclaw-ai/loongclaw --skill skill-creator-loongclaw-ai

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve?

It turns the messy process of inventing, documenting, and hardening Claude skills into a deliberate workflow so you can confidently launch and improve reusable agents for complex tasks.

Core Features & Use Cases

  • Intent capture and research: Interview the user, catalog tools and edge cases, and translate the answers into a structured skill draft.
  • Evaluation loop orchestration: Write test prompts, run with-skill and baseline evaluations, grade outputs, and aggregate benchmarks for each iteration.
  • Optimization and tooling: Bundle helper scripts, analyze metrics, document reviewer feedback, and fine-tune the SKILL.md description to improve triggering accuracy.

Quick Start

Ask Claude to summarize the desired workflow, draft the skill instructions, and design the evaluation prompts to run.

Frequently Asked Questions about skill-creator

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

FAQPage Schema
How do I automate Claude skill creation and evaluation?▼

You can automate Claude skill creation by running intent interviews, drafting SKILL.md files, and orchestrating iterative evaluation loops with test prompts to benchmark and refine performance.

What is the process for iterative improvement of Claude skills?▼

Iterative improvement involves writing test prompts, running with-skill and baseline evaluations, grading outputs, and aggregating benchmarks to fine-tune SKILL.md descriptions and helper scripts for reliable triggering.

How do I optimize Claude skill descriptions for better triggering accuracy?▼

You optimize skill descriptions by analyzing benchmarked evaluation metrics and reviewer feedback, allowing you to fine-tune the SKILL.md text so the skill triggers reliably for repeatable tasks.

Can I bundle helper scripts when building Claude skills?▼

Yes, you can bundle helper scripts alongside SKILL.md files and evaluation assets to ensure the skill instructions, tooling, and descriptions evolve together throughout the automation workflow.

Does this approach handle intent capture for new Claude workflows?▼

Yes, the intent capture mechanism interviews users to catalog tools and edge cases, translating the answers into a structured skill draft that captures new workflows for Claude.

What is the best way to benchmark Claude skills against a baseline?▼

The best way to benchmark Claude skills is to run parallel evaluations using test prompts against both a baseline and the active skill, then grade the outputs and aggregate the metrics for comparison.

Related Skills

Explore More Software Engineering Skills→
awood45awood45
11

capture-skill

Convert a conversation workflow into a structured Claude skill specification.

Community
Advanced
UnitOneAIUnitOneAI
44

model-supply-chain

Review AI/ML model supply chains for provenance, lineage, and backdoor indicators.

Official
Advanced
MCCTeamMCCTeam
2.1k

skill-creator

Create and iterate Claude skill drafts and evaluation workflows.

Official
Advanced
zhbonerzhboner

skill-creator

Draft SKILL.md files and refine them through prompt-based test evaluations.

Community
Advanced