skill-authoring

Create SKILL.md instruction documents with TOON decision trees and references.

1|Updated Jan 22, 2026
One-click install
npx skills add https://github.com/guardian-intelligence/apm2 --skill skill-authoring-guardian-intelligence
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: skill-authoring
Source: https://github.com/guardian-intelligence/apm2/tree/main/documents/skills/skill-authoring
Command: npx skills add https://github.com/guardian-intelligence/apm2 --skill skill-authoring-guardian-intelligence

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill provides a standardized framework for authoring agent instruction documents (SKILL.md) and their references, enabling consistent, auditable guidance creation.

Core Features & Use Cases

  • TOON-based decision-tree authoring: define instruction workflows that agents can follow deterministically.
  • Canonical Standards & Templates: reuse a central standard and templates for SKILL.md and references to ensure consistency across skills.
  • References Management: organize and link essential glossary and authoring guidelines to support rapid skill development, review, and compliance.

Quick Start

Initialize a new skill by creating SKILL.md at the root of the unit, reference the canonical standards in references/, and define a decision_tree entrypoint with STANDARD.

Frequently Asked Questions about skill-authoring

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

FAQPage Schema
How do I standardize agent instruction documents for consistent authoring?

Standardize agent instruction authoring by creating a SKILL.md file at the root unit, linking canonical standards in references, and defining a decision_tree entrypoint with STANDARD. This enforces frontmatter requirements and auditable guidance for consistent skill development.

What is a TOON decision tree for agent instruction design?

A TOON decision tree defines deterministic instruction workflows that agents can follow. It provides a standardized framework for authoring SKILL.md documents, ensuring consistent and auditable guidance creation for agent instruction design across skills.

How do I organize references for agent skill development and compliance?

Organize references by placing essential glossary and authoring guidelines in a references directory linked from the SKILL.md root. This supports rapid skill development, review, and compliance by reusing central canonical standards and templates.

Do I need frontmatter requirements when creating a SKILL.md file?

Yes, frontmatter requirements are mandated when creating a SKILL.md file. The skill enforces specific frontmatter fields including name and description to ensure consistent, auditable guidance for agent instruction design and governance.

What's the best way to structure decision trees for agent instruction workflows?

Structure decision trees using TOON-based authoring to define deterministic instruction workflows agents can follow. Initialize by creating SKILL.md at the root, reference canonical standards in references, and define a decision_tree entrypoint with STANDARD for consistency.

Can I reuse canonical standards and templates across multiple agent skills?

Yes, you can reuse a central standard and templates for SKILL.md and references across multiple skills. This ensures consistency in agent instruction design, enables rapid skill development, and supports governance through standardized authoring frameworks.