Skip to content

Chapter Metrics

Per-chapter content breakdown for the Custom Skill Developer guide, generated from a full scan of all 17 chapters.


Per-Chapter Breakdown

Chapter Title Word Count Code Examples Diagrams Tables Admonitions Links
1 What Are Claude Code Skills? ~1,390 3 1 1 2 1
2 The Skill Ecosystem ~2,010 9 1 4 5 0
3 Your First Skill ~4,100 18 1 3 6 1
4 SKILL.md Structure ~3,440 18 1 3 0 0
5 YAML Frontmatter ~2,280 22 0 3 0 0
6 Workflow Design ~4,740 31 2 4 0 0
7 Quality Scoring Systems ~3,170 16 0 6 0 0
8 Meta-Skill Routers ~2,890 10 0 2 0 0
9 Token Efficiency Design ~2,120 13 0 2 0 0
10 Session Logging ~2,580 11 0 6 0 0
11 Pipeline Orchestration ~2,610 10 1 3 0 0
12 Data Format Skills ~3,545 15 1 1 0 0
13 Code Generation Skills ~3,400 16 1 0 0 0
14 Analysis and Validation Skills ~3,290 9 2 2 0 0
15 Installation and Registry ~2,735 14 1 1 1 0
16 Testing and Debugging ~1,945 11 1 1 0 0
17 Publishing and Distribution ~3,225 20 0 3 0 0
Totals ~47,470 246 13 45 14 2

Column Definitions

Column Definition
Word Count Estimated from line count with content-density adjustment for mixed prose and code markdown
Code Examples Count of fenced code blocks (each opening+closing ``` pair = one block)
Diagrams Count of Mermaid diagram blocks (```mermaid opening fences)
Tables Count of distinct tables, identified by separator rows matching | --- | patterns
Admonitions Count of lines beginning with !!! (tip, note, warning, example callouts)
Links Count of inline external hyperlinks matching [text](https?://...) pattern

Notable Patterns

Most Code-Intensive Chapters

Rank Chapter Code Examples
1 Chapter 6 — Workflow Design 31
2 Chapter 5 — YAML Frontmatter 22
3 Chapter 17 — Publishing and Distribution 20
4 Chapter 3 — Your First Skill 18
4 Chapter 4 — SKILL.md Structure 18

Chapter 6 and 5 are the most code-dense because they cover the technical specification of skill structure — every rule is illustrated with a concrete YAML or markdown code block.

Most Table-Intensive Chapters

Rank Chapter Tables
1 Chapter 7 — Quality Scoring Systems 6
1 Chapter 10 — Session Logging 6
3 Chapter 2 — The Skill Ecosystem 4
3 Chapter 6 — Workflow Design 4

Chapters 7 and 10 lead in tables because they cover reference-heavy material — rubric criteria, scoring thresholds, and log schema specifications that are naturally expressed in tabular form.

Admonition Distribution

Admonitions (tip, warning, note, example callouts) are concentrated in the early foundational chapters:

Chapter Admonitions
Chapter 3 — Your First Skill 6
Chapter 2 — The Skill Ecosystem 5
Chapter 1 — What Are Claude Code Skills? 2
Chapter 15 — Installation and Registry 1

Chapters 1–3 use admonitions heavily to flag critical constraints for new readers (the 30-skill limit, session restart requirements, minimum viable skill format). Chapters 4–17 rely on structured prose and code blocks instead.

Longest Chapters by Word Count

Rank Chapter Word Count
1 Chapter 6 — Workflow Design ~4,740
2 Chapter 3 — Your First Skill ~4,100
3 Chapter 4 — SKILL.md Structure ~3,440
4 Chapter 13 — Code Generation Skills ~3,400
5 Chapter 12 — Data Format Skills ~3,545

Shortest Chapters by Word Count

Rank Chapter Word Count
1 Chapter 1 — What Are Claude Code Skills? ~1,390
2 Chapter 16 — Testing and Debugging ~1,945
3 Chapter 2 — The Skill Ecosystem ~2,010

Chapter 1 is intentionally lean — it establishes vocabulary and mental models without overwhelming new readers. Chapter 16 is concise because testing methodology is demonstrated through practical walkthroughs rather than extended explanation.


Part Summary

Part Chapters Total Words Total Code Examples Total Diagrams
Part I — Foundations 1–3 ~7,500 30 3
Part II — Skill Anatomy 4–7 ~13,630 87 3
Part III — Advanced Patterns 8–11 ~10,200 44 1
Part IV — Specialized Skills & Deployment 12–17 ~18,140 85 6

Part II (Skill Anatomy) is the most code-intensive section, covering the detailed technical specification of SKILL.md structure, YAML frontmatter, workflow design, and quality scoring. Part IV is the longest by word count, spanning six chapters across specialized skill types and the full deployment lifecycle.


How These Metrics Were Generated

This report was produced by the book-metrics-generator skill applied to the custom-skill-developer repository on 2026-02-20. All counts use the same methodology as book-metrics.md:

  • Word counts: Estimated from non-empty line counts with a content-density multiplier (~8 words per line for markdown with mixed prose and code)
  • Code blocks: Counted as pairs of triple-backtick delimiters; each pair = one block
  • Mermaid diagrams: Counted as occurrences of ```mermaid opening fences
  • Tables: Counted as separator rows matching | --- | patterns; each separator = one table
  • Admonitions: Counted as lines beginning with !!!
  • External links: Counted as occurrences of [text](https?://...) patterns within chapter files only