your recipes, without the life story.
the recipes you cook repeatedly are maybe thirty, and they are scattered: a screenshot, a book on a shelf, a website that now needs three dismissals before you can read it, and one your mother told you over the phone.
every recipe site is optimised for finding recipes, which is not the problem. the problem is reading the one you already know you are making, while cooking it.
a page with your thirty on it, typed plainly, loads instantly and stays readable with the screen at arm’s length. an agent will type them up for you if you paste them in.
a cookbook that is just the recipes.
01 / what it isthis is the one food recipe where the content lives in the page rather than in your browser, and that changes what it is for. because the recipes are in the html, anyone you send the url to can read them — so this is the page you share with a sibling, and it is also the page you should not put anything private in.
adding a recipe means asking your assistant to add it and republishing, which sounds worse than it is: you paste the recipe into the conversation and it gets typed up properly, with the ingredients separated from the method and the story removed.
- your recipes, ingredients and method, typed properly
- scaling — double it, halve it, without arithmetic
- an index and a search that work on thirty recipes
- a cooking view with big text and no distractions
- print styles, because some people want it on paper
- a page that opens in under a second on a kitchen phone
- where the recipes live
- in the published page. readable by anyone with the url
- adding one
- paste it to your assistant, republish to the same url
- scaling
- computed in the page from the quantities as written
- what it stores
- nothing per-visitor, unless you asked for notes — and then that part is local
see one in action.
02 / examplefourteen household recipes, ingredients in one column and method in the other, with scaling
the live example is still being built — the prompt below is the one it was made with.
copy this, then make it yours.
03 / starting prompta starting point, not a template — the point of an agent is that the result is yours. change the sections to match what you want, and remove anything you don’t need.
Build me a recipe collection as a single web page. The recipes go in the page itself.
Ask me first:
- Which recipes to start with. I will paste them in — they may be messy, copied from websites, or dictated.
- Whether I cook in metric, imperial, or both.
- Whether I want scaling, and what the common cases are (double, halve, per person).
- Whether I want a notes area per recipe for what I changed last time.
Type up exactly what I give you. Do not invent recipes, do not adjust quantities, do not substitute ingredients, and do not add a recipe I did not supply. If a recipe I paste is ambiguous, ask rather than guessing — a wrong quantity here ruins dinner.
WHAT IT DOES
- An index of every recipe, and a search that filters it as I type.
- Each recipe: title, how long it takes, how many it serves, ingredients as a list, method as numbered steps. Nothing else unless I asked for it.
- Scaling: change the number of servings and the quantities update. Handle fractions sensibly and leave anything unscalable (a tin, an oven temperature, "a pinch") alone.
- A cooking view: one recipe, large type, ingredients and method both visible if the screen allows, no navigation in the way.
- Print styles that produce a clean one-page recipe.
- If I asked for notes: a per-recipe notes area saved in my browser, clearly marked as local to this device and not part of what other people see.
WHAT IT MUST NOT HAVE
- No introductory story, no headnote I did not write, no photographs unless I supply them.
- No nutrition information — you would have to invent it.
- No links out to recipe sites.
DESIGN
- Readable from a metre away with the phone propped up. Large type in the cooking view is the single most important design decision here.
- Fast: no web fonts blocking text, no images unless I gave you some.
- Sensible on a laptop for reading and on paper when printed.
BE HONEST ON THE PAGE
- The recipes are in the page, so anyone with the url can read them. If I asked for local notes, say clearly that those are stored in this browser only and are not shared.
DELIVERY
- Produce one self-contained HTML file. Inline the CSS and JavaScript.
- No build step, no server, no framework, no package install, no external
requests at runtime.
- Mobile-first and responsive; it has to be usable one-handed on a phone and
still look right on a laptop.
- Accessible: semantic HTML, real contrast, visible keyboard focus, labelled
controls, touch targets big enough to hit, and respect reduced-motion.
- Use placeholders where information is missing. Do not invent facts, numbers,
names, dates, prices, or results I have not given you.
- Before you call it finished, use it yourself: enter real-looking data, reload
the page, and confirm nothing is lost or mangled.
PUBLISHING
Publish the finished page to HTMLShare and give me the link.
Use exactly one of the two paths below, preferring the first. Make the call
directly. Do not route it through a code sandbox, container, or computer-use
session — that is where the request gets blocked, and it is never necessary.
1. An HTMLShare integration or connector, if you have one configured.
The page is permanent and owned by my account. Report the url and the
siteId. There is no claimUrl or manageToken on this path — do not go
looking for one, and do not switch paths in order to produce one.
2. Otherwise, publish anonymously over REST. No account, key, or signup is
needed:
POST https://api.htmlshare.net/v1/publish
content-type: application/json
{"html": "<!doctype html> ... "}
The reference is https://htmlshare.net/docs and, for the full schema,
https://api.htmlshare.net/openapi.json
Report these from the response, in your reply and never inside the page:
url the public page
claimUrl a private link that is mine alone. opening it is how I keep
the page past its 24-hour expiry, or set an email reminder
before it goes. it carries the manage token, so it must
never appear in the published HTML or anywhere a visitor
could see it.
manageToken the only thing that can edit or delete this page later. it is
returned once and never again, so tell me what it is.
Tell me which of the two you used. If the path you reach for first is
unavailable in your environment, say so and tell me what you are switching to
before you publish — a permanent page on my account and a 24-hour anonymous
page are different things, and which one I get is my decision, not a detail.
If I ask for a change afterwards, update the existing page instead of
publishing again, so the URL I have already shared keeps working — through the
same integration on path 1, or with the manage token as an X-Manage-Token
header on path 2.
RECIPE MARKER
Include this tag in the <head>, exactly as written:
<meta name="htmlshare-recipe" content="recipe-collection">
It tells htmlshare which recipe the page came from, so the library can count
that this one was actually made. It carries nothing else.give it a url.
04 / publishing ithow you publish depends on how you’re working.
- with an integration
- if your assistant already has htmlshare connected — mcp or another supported integration — just ask it to publish the finished page and hand back the url. the page belongs to your account and is permanent. when you want a change, ask for an update to that same page rather than a new one.
- with no integration
- the prompt tells the assistant to post the html to the rest api instead, which needs no account and no key. it comes back with a live url you can open immediately, a private claim link, and a manage token. the claim link is how the page stops being temporary — keep it.
- by hand
- nothing stops you doing it yourself: have the assistant give you the html file and drop it on the htmlshare home page. same url, same page, one more step.
- what matters here
- this page grows by being republished, which makes updating in place the whole workflow: same url, new recipe, and the link you sent your sister still works. claim it and keep the manage token where your assistant can use it.
a recipe book is a decade-long object.
05 / keeping it aroundfine for typing up one recipe to see the format. the page stops serving 24 hours later, with a 7-day recovery window behind that.
claim it. this is a page you add to for years and send to people, and both of those need the address to be permanent.
- keep it around
- a recipe collection is a years-long object and an anonymous page lasts a day.
- stable updates
- you add recipes by republishing. the link you gave your family has to keep pointing at the current version.
- custom url
- this one gets shared out loud — a slug someone can type from hearing it is worth having.
questions.
06 / questionscan other people see my recipes?
yes, if they have the url — the recipes are in the page, which is what makes this the one food page you can genuinely share. do not put anything in it you would not hand out.
how do i add a recipe later?
paste it to your assistant and ask for it to be added, then republish over the same url. the manage token from the original publish, or your account, is what authorises that — so the link everyone has keeps working.
will it scale quantities correctly?
for anything numeric, yes, and the prompt tells it to leave alone the things that do not scale — oven temperatures, a tin of tomatoes, a pinch. check the first one you scale, since this is where hand-built pages tend to be slightly wrong.
can i import from a recipe website?
the page cannot fetch anything, but your assistant can read what you paste. copying the text into the conversation and asking for it to be typed up is the workflow, and it strips the story out on the way through.