{} Claude Skills · ch.11 · examples gallery
📚 Chapter 11 · The canon

Six skills worth reading

After ten chapters of theory, here's the canon: six real skills shipped today, ranged from the smallest viable one to the largest orchestrator. Read them like recipes — each teaches a different lesson about the form.

You've seen the theory: descriptions are triggers, SKILL.md should be short, supporting files do the heavy lifting. Now read the actual skills. Six of them, picked to span the form: from a fifty-line one-trick pony to a multi-file orchestrator that spawns agents.

For each one we'll write a short paragraph, then call out the lesson it teaches that the others don't. At the end we'll name the five patterns visible in all six.

1The canon, one at a time

Each skill below is real — they all ship in the user's ~/.claude/skills/ today. They're ordered by complexity, smallest first.

📚
You've seen six skills now. The next section is the part you'll quote back: the patterns visible in all of them. They're not coincidences — they're what makes the format work.

2Patterns visible in all six

Six skills, ranged from fifty lines to three hundred, single-script to multi-agent orchestrator. They look different on the surface. Underneath, they all share five things.

  • All have specific trigger phrases — not "anything related to X". Concrete verbs and nouns: "grab the transcript", "verify a PR", "build a book about". Claude reads the description and matches against real user phrasing.
  • All have a clear "when to use" boundary — either as an explicit list ("when the user wants to start a new book"), or as a SKIP clause that disqualifies confusing neighbours (the claude-api skill literally says "SKIP when code imports openai").
  • All keep SKILL.md scannable — even the longest ones use short paragraphs, clear headings, and link to reference files for deep content. You can read any of them top-to-bottom in under three minutes.
  • None mix unrelated concernsverify doesn't also write reports. update-config doesn't also run hooks. book-builder doesn't also do code reviews. If you'd want it as a separate verb, it's a separate skill.
  • Each does one job well — and the test for that is: would you describe it in a single sentence to a colleague? If you can't, it's two skills pretending to be one.

If a skill of yours doesn't have all five, it's not necessarily broken — but it's probably the reason something feels off when you invoke it.

3Browse the canon

Click any of the six cards to expand a detail panel: actual description excerpt, folder tree, when to study it, and the three lessons it teaches. Default state is collapsed — pick the one that looks closest to what you're trying to build.

Interactive · gallery browser Click a card · expand the detail panel

4Use-case matcher

Type what you want to build. The matcher compares your description against the six gallery skills' lessons and suggests which one or two are closest in shape. Try the example chips first to see how it picks.

Interactive · use-case matcher Describe a skill you want · see what it resembles
Type a description above (or click an example) to see what it resembles.

5Anatomy comparison

Pick two or three skills to compare side-by-side: their frontmatter description, body length in lines, the supporting files they ship with, and the one-line "signature pattern" each one represents.

Interactive · anatomy comparison Pick 2 or 3 skills · compare side by side

6Two real SKILL.md files

The first tab is an excerpt of the book-builder SKILL.md — the largest orchestrator in the canon. The second is the full text of youtube-transcript — the smallest viable one. Read them side by side to see the spread.

The shapes are different — one routes work to templates, the other defines the work inline — but the patterns are the same. Specific triggers, clear "when to use" boundary, scannable body, one job each.