The magic powers of HTML

Ask an AI agent for a document and you often get a markdown file or a long response in the chat about the thing. Ask for an HTML and you can get the thing itself, one you move, walk through and keep. The words you add decide which.

Every AI agent that can write a file can write an HTML file, and the tool you use decides where it lands. This guide is about the powers themselves, and about the fact that every one of them has a plain name you can ask for.

# Q4 in one page

Revenue **up 12 percent**
- Now
- Proposed

<h1>Q4 in one page</h1>
<input type="range">
<button>Compare</button>

The powers were always there

An AI agent will give you what you ask for. The challenge is that most people do not know what to ask for.

Most days I ask for the same thing without noticing, and what comes back is fine. The difference between an answer you skim and a thing you keep is often the last three words of the prompt.

The phrases are things like add sliders, make it a wizard, make it interactive and not so overwhelming. Plain words, the kind you would say to a colleague.

Why the words work at all

Thariq Shihipar, who works on Claude Code at Anthropic, made this case in The unreasonable effectiveness of HTML, and his page of examples argues it better than another paragraph from me would. What he leaves open is which words to type, and that is the rest of this guide.

The agent defaults to prose because prose is what it is asked for. The words at the end of the request are what tell it to build something else.

Seven powers, and what to say

Let me move it. Add sliders, add a filter, let me change the assumptions. A model of anything contains numbers somebody guessed. Written as prose, those guesses are invisible and the reader argues with the conclusion. Put them on sliders and the reader argues with the guesses instead, which is the argument worth having.

Take me through it. Make it a wizard, one step at a time, ask me the questions first. Some things are not documents, they are decisions with an order. A page that asks you one question, then the next question based on your answer, is doing work that a list of all the questions at once cannot do.

HTML is great for storytelling

Some points need to arrive one at a time. A document cannot hold the second idea back until the first has landed. A page can, because the browser knows how far you have scrolled. The long reads newspapers build this way, with charts that fill in as you scroll, are HTML and nothing more.

Walk me through the point. Make it a scrolling story, reveal it one step at a time, let the figure draw itself as I scroll. A wizard waits for the reader to answer. A story does not wait for anything, it just refuses to arrive all at once. The argument builds as you scroll, one figure at a time, and the reader is never holding more than one idea. We build these for our own guides, and the thing that makes them work is not the animation. It is that somebody had to decide the order.

An example. How does the context window in AI agents work?

That is one file. The glass, the eight scenes and the scrolling are all in it, and it opens on any laptop with nothing installed.

Four more things to ask for

Show me less. Not so overwhelming, one screen, put the detail behind a toggle. Agents are generous, and generosity reads as noise. This is the phrase I reach for last and probably should reach for first, because asking for less feels like asking for worse. What it actually asks for is a hierarchy.

Show me more than one. Give me three directions, side by side. A single answer invites a yes or a no. Three answers next to each other invite a choice, and a choice is where taste gets applied. The three do not have to be equally good. Two of them exist so that the third has something to be better than.

Make me want to look at it. Make it magical, make it feel alive. This is the vaguest phrase here and it works anyway, which is worth being curious about rather than embarrassed by. The output of a vague instruction is motion, weight and a sense that somebody cared, and those are the things people actually mean when they say a thing feels finished.

Let me keep it. One file I can send, make it printable. A page somebody cannot open is a draft. One file with nothing to install is the difference between a thing you show and a thing that travels.

Interactive or alive

I use two of these as though they were the same instruction, and I am starting to think they pull in different directions.

Make it interactive asks for controls. I get buttons, tabs and inputs, sometimes more of them than the idea needed.

Make it feel alive asks for response. I get motion, timing and state, usually with fewer controls rather than more.

The next time you ask for interactive and get a control panel you did not want, the second phrase may be the one you were reaching for.

One version, then shape it

Do not try to say everything in the first request. Ask for the page, look at it, and say what is off. Each round is a few more words, and each round lifts what comes back.

The first version is plain and a bit generous. Add sliders puts the numbers in your hands. Make it a wizard turns the list into steps. Not so overwhelming takes away what you did not need. The page you end up with is shaped by your feedback, one round at a time.

When you have the page you want, ask the AI agent to turn it into a skill. The rounds you just went through become the starting point next time, and the first version arrives where this one ended.

We say the same thing in our training. Our contribution is taste, and taste is applied round by round. Put the guesses on sliders and you find out which ones you believe. Walk through the decision and you notice the step you skipped. You are checking your own thinking, on a page you can point at.

What to do tomorrow

Take one thing you asked an agent for this week, something that came back as text and was fine, and ask for it again with three more words on the end. The same request with an ending, nothing more.

If you find yourself typing the same ending every time, that is worth noticing. A phrase you repeat is a preference, and a preference you write down once is a skill.

The clearest example I know comes from Thariq Shihipar again. In a post on X he shared the skill people at Anthropic had been using a lot, ELI5, which you call with /eli5 and whatever you want explained. The whole skill rests on one sentence.

explain like I'm someone who knows nothing about this topic, using a HTML artifact with big pictures and few words

Now go make some awesome HTMLs!