Tutorial

Tutorial Writer

Tutorials fail when they front-load theory or hide the result until the end. This structure builds toward a visible goal with checkpoints that confirm understanding along the way.

Structure (non-negotiable order)

1. Show the End State First

Before any steps, show what the reader will have built when they're done. A screenshot, a code snippet output, a running demo. This creates a goal — readers tolerate confusion when they can see where it's going.

By the end of this tutorial, you'll have:
[concrete, visible result — not "understand X" but "a working Y that does Z"]

2. Prerequisites

What the reader needs to already know or have installed. Be specific:

• "Node.js 18+" not "Node.js"
• "Familiarity with async/await" not "JavaScript experience"
• Link to setup instructions for tools, don't include them inline

3. Steps in Order of Dependency

Each step does one thing and builds on the previous step. Never skip a step because it "should be obvious" — if it were obvious, there would be no need for a tutorial.

Each step structure:

## Step N: [Action verb + what]

[1-2 sentences of context — why this step exists]

[Code block or command to run]

[Expected output or visible result]

4. Working Checkpoint After Each Major Step

Every 2-4 steps, include a checkpoint: the reader should run something and see a result that confirms they're on track. If they can't, they know exactly where they got lost.

### Checkpoint
Run:
[command]

You should see:
[exact expected output]

If you see X instead, [specific fix].

5. "What Just Happened?" After Each Checkpoint

A brief explanation of the mechanism — not re-explaining the steps, but the underlying concept. This is what turns a recipe-follower into someone who understands the system.

Keep it short: 3-5 sentences. Long explanations belong in concept docs, not tutorials.

6. Extension Exercises

After the tutorial is complete, 2-3 things to try next. These extend the tutorial naturally rather than pointing at unrelated docs. Format: "Try X to see what happens when Y."

Writing Rules

Every code block must:

• Be copy-pasteable without modification (no `` placeholders without explaining how to get the value)
• Show exactly what to run and where (file path matters)
• Include the full context needed to understand it, not just the interesting line

Errors the reader will hit: Include the 2-3 most common errors with exact error messages and fixes. Don't ignore the unhappy path — it's where most readers abandon the tutorial.

Words to never use:

• "simple" / "simply" / "easy" / "just" — if it were, they wouldn't need a tutorial
• "obviously" / "of course" — condescending
• "as you can see" — if the code needs narrating, the code isn't clear

Workflow

1. Clarify the topic and target audience (what do they already know?)
2. Identify the end state — what working thing will they have built?
3. Map the dependency order of steps before writing any of them
4. Write with checkpoints every 2-4 steps
5. Test each code block mentally or actually run it — tutorials with broken code destroy trust immediately

Output

A complete tutorial with:

• End state shown upfront
• Prerequisites list
• Numbered steps with code blocks
• Working checkpoints with expected output
• "What just happened?" after each checkpoint
• Common errors section
• Extension exercises

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: kriscard
• Source: kriscard/Skills
• License: MIT

Install and usage instructions live in the source repository linked above.