Back to list
chkim-su

skill-design

by chkim-su

1🍴 0📅 Jan 16, 2026

SKILL.md


name: skill-design description: Best practices for skill structure and types. Use when creating skills. allowed-tools: ["Read", "Write", "Glob", "Grep", "Bash"]

Skill Anatomy

skill-name/
├── SKILL.md         # Required: core instructions (<500 words)
├── scripts/         # Optional: executable code
├── references/      # Optional: detailed docs (load on-demand)
└── assets/          # Optional: templates, images

Skill Types

TypeFreedomUse whenStructure
KnowledgeHighMultiple approaches, context-dependentSKILL.md + references/
HybridMediumGuidance + scripts neededSKILL.md + scripts/ + references/
ToolLowDeterministic, repeatable opsSKILL.md + scripts/
ExpertVery LowComplex internals, undocumented APIsFull + validation/

When to Script

  • Same code rewritten repeatedly → Script
  • File format manipulation → Script
  • Reliability critical → Script
  • External API/tool integration → Script

When to Hook (Enforcement Logic)

Rule: If behavior MUST happen, use hooks. Documentation cannot enforce.

Requirement TypeExampleSolution
MUST/REQUIRED"MUST use Skill tool"PreToolUse hook to warn/block
Validation"Schema must be valid"PostToolUse hook to validate
Prevention"Never commit secrets"PreToolUse hook to block
Guidance"Consider using X"Documentation only (OK)

Hook vs Documentation Decision

If keyword in ["MUST", "REQUIRED", "CRITICAL", "강제", "반드시"]:
    → Implement as Hook (documentation alone WILL be ignored)

If keyword in ["should", "consider", "recommend"]:
    → Documentation is sufficient

Hookify Checklist

Before finalizing skill:

  1. Search for enforcement keywords (MUST, REQUIRED, CRITICAL)
  2. Each enforcement requirement → corresponding hook exists?
  3. If no hook → either create hook OR downgrade to "should"

Example:

  • ❌ "MUST use Skill() tool" in SKILL.md → Agents ignore this
  • ✅ PreToolUse hook on Read/Grep/Glob warns when skill files accessed directly

SKILL.md Structure

---
name: skill-name
description: What + when to use (triggers skill loading)
allowed-tools: ["Tool1", "Tool2"]
---
# Skill Name

[2-3 sentence overview]

## Quick Start
[Fastest path]

## Workflow
1. Step 1
2. Step 2

## Scripts
| Script | Purpose | Usage |
|--------|---------|-------|

## Key Principles
- Principle 1
- Principle 2

For advanced: [references/advanced.md]

Tool Restrictions

# Knowledge (read-only)
allowed-tools: ["Read", "Grep", "Glob"]

# Hybrid (guidance + generation)
allowed-tools: ["Read", "Write", "Grep", "Glob", "Bash"]

# Tool (file manipulation)
allowed-tools: ["Read", "Write", "Bash"]

Checklist

  • SKILL.md < 500 words
  • Description includes trigger phrases
  • Scripts tested
  • References linked
  • allowed-tools matches purpose
  • Hookify check: Enforcement keywords (MUST/REQUIRED/CRITICAL) → hooks exist

Score

Total Score

60/100

Based on repository quality metrics

SKILL.md

SKILL.mdファイルが含まれている

+20
LICENSE

ライセンスが設定されている

+10
説明文

100文字以上の説明がある

0/10
人気

GitHub Stars 100以上

0/15
最近の活動

3ヶ月以内に更新がある

0/10
フォーク

10回以上フォークされている

0/5
Issue管理

オープンIssueが50未満

+5
言語

プログラミング言語が設定されている

+5
タグ

1つ以上のタグが設定されている

0/5

Reviews

💬

Reviews coming soon