skills.rest
terzigolu
terzigoluCOMMUNITY
Explore all skills by this author →
GitHub

create-skill

Automate drafting, evaluating, and iterating AI skills with structured prompts and metrics.

37|1|Updated Apr 28, 2026
Search Tags:#workflow#evaluation#iteration#creation#benchmark#skill
One-click install
npx skills add https://github.com/terzigolu/warp-lite --skill create-skill-terzigolu
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: create-skill
Source: https://github.com/terzigolu/warp-lite/tree/main/resources/bundled/skills/create-skill
Command: npx skills add https://github.com/terzigolu/warp-lite --skill create-skill-terzigolu

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

This Skill guides teams through the end-to-end process of creating, testing, and improving AI skills. It provides a repeatable workflow to draft SKILL.md, design evaluation prompts, run benchmarks, and iterate based on results.

Core Features & Use Cases

  • Draft-to-eval loop: generate initial skill drafts, link test prompts, and run evaluations to measure triggering accuracy.
  • Evaluation orchestration: collect results, compute metrics, and surface actionable insights to guide improvements.
  • Iterative refinement: apply feedback from evals to rewrite descriptions, adjust prompts, and re-run tests until goals are met.
  • Packaging and deployment prep: prepare a final skill package and optional metadata for distribution.

Quick Start

Define a new skill in SKILL.md, create an initial eval set, run the evaluation loop, review results, and iteratively improve until you’re satisfied.

Frequently Asked Questions about create-skill

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

FAQPage Schema
How do I create and test AI skills using a structured workflow?▼

To create AI skills, you draft a SKILL.md file, design evaluation prompts, run benchmarks to measure triggering accuracy, and iteratively refine descriptions based on results.

What is the best way to evaluate and refine prompts for AI skills?▼

The best way to evaluate prompts is running a draft-to-eval loop where you link test prompts, compute metrics, and apply feedback to rewrite descriptions until performance goals are met.

Do I need yaml to automate skill creation and benchmarking?▼

Yes, yaml is required as a dependency to support the repeatable workflow that automates skill creation, evaluation orchestration, and optional packaging for distribution.

Can I prepare a skill package for distribution after iterating on benchmarks?▼

Yes, after you iteratively refine your skill through evaluation loops, you can prepare a final skill package and optional metadata specifically formatted for distribution.

Why does my AI skill trigger inaccurately during evaluation?▼

Inaccurate triggering during evaluation indicates a need for iterative refinement, where you apply feedback from computed metrics to rewrite descriptions and adjust prompts.

What is a draft-to-eval loop for building AI skills?▼

A draft-to-eval loop is a repeatable workflow that generates initial skill drafts, links test prompts, runs evaluations to collect metrics, and surfaces insights to guide improvements.

Related Skills

Explore More Product & Management Skills→
yu13130122297yu13130122297

scout

Identify uncertainties in research framing, datasets, metrics, and baselines.

Community
Intermediate
vnmoorthyvnmoorthy
2

canary

Monitor live apps for console errors and performance regressions post-deploy.

Community
Advanced
alex-robert-fralex-robert-fr

skill-creator

Automate skill creation, iteration, and benchmarking with SKILL.md generation and evaluation loops.

Community
Advanced
mneves75mneves75
74

writing-skills

Apply Test-Driven Development to create and test AI skills.

Community
Intermediate