skills.rest
Sailfishc
SailfishcCOMMUNITY
Explore all skills by this author →
Website𝕏@sailfishcc1GitHub

creating-agent-skills

Design, package, validate, and evaluate reusable agent skills across Codex and Claude workflows.

Updated Apr 25, 2026
Search Tags:#validation#codex#packaging#claude-code#agent-skills#skill-design#eval-cases
One-click install
npx skills add https://github.com/Sailfishc/guanfu --skill creating-agent-skills-sailfishc
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: creating-agent-skills
Source: https://github.com/Sailfishc/guanfu/tree/main/skills/creating-agent-skills
Command: npx skills add https://github.com/Sailfishc/guanfu --skill creating-agent-skills-sailfishc

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Enables teams to design, package, validate, and evaluate reusable agent skills across Codex, Claude Code, Claude.ai, and API deployments. Systematizes the creation process to reduce drift and accelerate skill evolution.

Core Features & Use Cases

  • Eval-driven skill authoring scaffolds.
  • Frontmatter, references templates, and packaging guidelines for cross-platform skills.
  • Guidance for PKG structure, mirroring to Codex/Claude, and evaluation case design.

Quick Start

Create a new skill by following the guidance in this SKILL.md, including writing the SKILL.md frontmatter and the evaluation scaffolding.

Frequently Asked Questions about creating-agent-skills

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

FAQPage Schema
How do I design and validate reusable agent skills for Claude Code?▼

Reusable agent skills are designed and validated using frontmatter conventions, templates, and packaging guidelines. Static checks enforce frontmatter, references, and package structure to ensure skills are ready for Claude Code distribution.

What is the best way to package agent skills across Codex and Claude.ai deployments?▼

Packaging agent skills for Codex and Claude.ai requires following specific package structure guidelines and mirroring workflows. This systematizes the creation process to reduce drift and accelerate skill evolution across platforms.

How do I write evaluation cases for agent skills?▼

Evaluation cases for agent skills are written using eval-driven authoring scaffolds. This provides structured guidance for evaluation case design to validate skill behavior before distribution.

Does this skill authoring workflow support API deployments?▼

Yes, the skill authoring workflow supports API deployments along with Codex, Claude Code, and Claude.ai. It provides frontmatter conventions and packaging guidance to ensure ready-to-distribute skills across these platforms.

Why do I need static checks for agent skill frontmatter and package structure?▼

Static checks for frontmatter and package structure are needed to prevent distribution errors and reduce drift. Enforcing these checks ensures that references and package structure meet ready-to-distribute standards.

Related Skills

Explore More Software Engineering Skills→
nirukk52nirukk52
1

skill-creator

Create and package Claude skills with SKILL.md templates and validation.

Community
Advanced
aws-samplesaws-samples
21

scaffold

Generate runnable agent project scaffolds from basic, multi-agent, RAG, and tool templates.

Official
Advanced
metabenchmetabench

static-analysis

Validate vanilla JS package code with linters, type checkers, and AST-driven tools.

Community
Intermediate
ClawStackStudiosClawStackStudios
19

@gitlobster/bridge

Migrate legacy OpenClaw skills into GitLobster packages with manifest.json and SKILL.md.

Community
Basic