skill-authoring

Guide authors in creating cross-platform, agent-agnostic skills with a three-layer architecture.

Updated Aug 23, 2026
One-click install
npx skills add https://github.com/contextware/skills --skill skill-authoring-contextware
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: skill-authoring
Source: https://github.com/contextware/skills/tree/main/skills/skill-authoring
Command: npx skills add https://github.com/contextware/skills --skill skill-authoring-contextware

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

This Skill provides a comprehensive guide and best practices for creating AI agent skills that are portable across different platforms and agents, ensuring consistency and maintainability.

Core Features & Use Cases

  • Guidance on Skill Structure: Learn how to organize SKILL.md files, scripts, and references.
  • Portable Workflow Design: Understand how to write instructions that agents can interpret and execute regardless of their underlying platform.
  • MCP Server Integration: Learn how to declaratively define MCP server requirements and provide reference URLs for broader compatibility.
  • Use Case: When developing a new skill for your AI agent, use this Skill to ensure it adheres to best practices, making it usable by other agents and platforms.

Quick Start

Use the skill-authoring skill to review the guidelines for creating a new portable skill.

Frequently Asked Questions about skill-authoring

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

FAQPage Schema
How do I author AI agent skills that work across different platforms?

To author portable AI agent skills, use a three-layer architecture separating a portable layer from platform adapters. This approach ensures your natural language instructions and declarative contracts execute consistently across different agents.

What is the best way to structure MCP server requirements in an AI skill?

The best way to structure MCP server requirements is by using frontmatter to declaratively define them and providing reference URLs. This ensures broader compatibility and clear contracts for the agent to interpret.

How do I handle user-blocking actions when writing agent scripting guidelines?

Handle user-blocking actions by implementing specific design patterns provided in skill authoring guidelines. These patterns ensure your self-contained scripts manage execution pauses and user confirmations safely.

Can I write natural language instructions that any AI agent can interpret and execute?

Yes, you can write natural language instructions for any AI agent by adhering to declarative contracts and a portable layer architecture. This makes your skills agent-agnostic and maintains cross-platform consistency.

Why does my portable skill fail to execute consistently across different platforms?

Portable skills fail across platforms when they lack proper platform adapter layers or self-contained scripts. Troubleshoot by verifying your frontmatter requirements and ensuring your declarative contracts are correctly structured.