llms.txt
# Your brand
Products: /products
Policies: /shipping, /returns
Trust: /endorsed.jsonllms.txt. A simple file on your site that tells AI systems who you are, what you sell, and where to find the important details.
Your next customer might not be a person.
Be the brand AI shoppers can trust.
Good for Humans makes your products readable and your claims verified, so shopping agents can find you, check you, and choose you.
For shopping agents. Humans welcome.
GOOD FOR HUMANS / INDEXLIVE
headline
Be the brand AI shoppers can trust.
description
Good for Humans makes your products readable and your claims verified, so shopping agents can find you, check you, and choose you.
action.target
#assessment
The shift: AI went from searching to choosing
For years, being found online meant showing up in search results. A person typed a question, scanned a page of links, and clicked.
AI agents now take a request like "find me a linen dress for a summer wedding, under $200, that I can return easily" and do the work. They search, filter, compare, and come back with a pick. Sometimes they come back ready to buy.
When a shopper hands the job to an agent, the agent decides which products they see and which sellers make the cut. Your customer may never visit a page of results at all.
That makes the agent a new kind of gatekeeper. And it has its own way of deciding.
How an AI agent decides what to buy
Shopping experiences differ by platform. The following is an illustrative decision flow, not a description of every AI agent or a report of our own tests.
It starts with what it knows about the shopper. Size, preferences, past conversations. Your product has to fit a specific person, not a general audience.
It may search connected product catalogs or browse merchant sites, depending on the platform. Catalogs can organize products by name, photo, price, and availability. If you are hard to read, you may never make that short list.
It filters before it ranks. Bad matches get dropped. It judges whether a product is trustworthy. It tries to confirm the item is in stock.
An agent may check third-party sources, like review roundups, Reddit and forums, alongside connected catalogs and merchant sites. The sources it uses and how it weighs trust signals depend on the platform.
The signals that matter most
Put it together and a clear picture forms. To be chosen, your brand needs:
Structured product data. Sizes, materials, fit, care, ingredients, colors, and who the product is for, in a form a machine can read without guessing.
Clear pricing, with all fees. The full price, shipping, taxes, and any recurring charges, shown plainly before commitment.
Verifiable claims and certifications. "Organic," "fair trade," "cruelty free," "clean." If you say it, an agent wants to see proof.
Reviews and reputation. What real customers say, and whether you are a seller the agent can trust.
Availability. Accurate, current stock, by size and shade.
An easy way to complete the purchase. If the agent can't finish checkout, it may pick someone it can.
Clear returns and refunds. Simple, honest policies an agent can read and explain.
Overall trustworthiness. Consistency between what you say, what you show, and what others say about you.
For service providers
For services, make your credentials, service area, availability, pricing, policies, and booking options clear and checkable.
This website has two audiences. You are one of them.
See the same page the way an AI agent sees it.
See what agents seeIs this the same as SEO?
SEO helps people find and click your pages. Agent readiness builds on that foundation with machine readable facts, verifiable claims, and a way to complete a purchase.
What happens to companies and brands that ignore this
But a growing share of shopping starts with a request to an agent. And when an agent can't read your sizing, can't confirm your stock, can't verify your "clean" claim, can't verify the services you offer or can't complete checkout, it doesn't call you to ask. It quietly moves on.
For values driven companies, there is an extra cost. If your sustainability, sourcing, or ingredient standards aren't readable and provable, an agent can't tell you apart from a brand that just uses the same words.
What agent ready looks like
Agent readiness is not one tool. It is a set of clear, honest signals that work together and are updated regularly.
schema.json
{
"@context": "https://schema.org",
"@type": "Product",
"name": "Your product"
}Schema and structured data. Product, price, availability, review, and organization markup that turns your pages into facts a machine can read.
endorsed.json
{
"brand": "Your brand",
"verified_claims": [],
"evidence": []
}A trust file, such as endorsed.json. A machine readable record of your verified claims and certifications, so an agent can check what you stand for instead of taking your word for it.
How Good for Humans helps
We make you findable and readable by AI Agents, your new customers.
The Good for Humans seal
The Good for Humans seal is an agent readable trust seal. Alongside it, we publish a clean, simple human page explaining the same verified facts. Same facts, two audiences.
Verification of your claims
A trust seal only means something if it is earned. We review the claims you make, such as materials, sourcing, ingredients, and certifications, and confirm what can be backed up. What goes into your seal is what we have checked.
Agent Readiness Assessment
We review your company and website for readable product details, full prices, verifiable claims, clear policies, and purchase access. You get a clear picture of what to fix first.
We build this for you on your website
Some fixes are simple. Others take real work: product feeds, structured data, policies, and connectors that let agents complete a purchase. Good for Humans is a project of The Tin Shed Company LLC. Tin Shed can help you build what your assessment recommends.
Every file we build for brands lives on this site too.
Who it's for
Online retail brands, first. You built your business online, for a customer who shops on a screen. Now that customer is sending an agent to shop for them. Sizing, materials, pricing, stock, and returns are the details agents check first. Make them machine readable and you give agents every reason to recommend you.
Service providers, next. Mortgage brokers, dentists, lawyers, roofers, plumbers: anyone people have to find, research, and vet before they hire. That research is moving to agents, who now compare credentials, reviews, pricing, and availability for their humans. The providers who are easiest to understand and verify are the ones who get the call.
Values-driven companies of every kind. If you make honest products, provide real value, treat people well, and stand behind your work, we want agents to see it, trust it, and send people your way.
How it works
1. Assess.
We review your company and your current website through an agent's eyes and show you what helps, what hurts, and what is missing.
2. Publish.
We help you put it all in place on your site and in your product feeds.
3. Seal.
We verify your claims and build your Good for Humans seal: llms.txt, endorsed.json, and structured data, plus your human page.
4. Monitor.
Agents and their rules change. We keep an eye on how agents see your brand and help you keep your signals accurate and current.
Frequently asked questions
Be the company the agent can trust
Start with an Agent Readiness Assessment. We will show you how AI agents see your brand today, and what it will take to be chosen tomorrow.