Your agent/ChatGPT

Let ChatGPT hand you a link, not a file.

It will happily build you the page, the tool, the little app. Add agenthost as a connector and it can put the thing online too, at an address that works on anyone’s phone.

A download is not a thing you can send someone.

Ask ChatGPT for a small tool and it will write you one. Then it offers you a file, and the file is where it ends. You cannot email it to a colleague in any way that works, you cannot open it on your phone, and next week you will not remember which folder it went in.

agenthost is the missing half. Once it is connected, ChatGPT can send those files to us and get back a working https address instead. There is nothing for either of you to configure: no build step, no framework to declare, no server to pick. It reads the files, works out what they are, and runs them.

It works for the plain page and for the thing that has to actually run: a Python script behind a form, a small Node API, something that needs to remember what people typed. And you never handle a key, because signing in happens in your browser rather than in the chat.

You ask for it in a sentence.

There are no commands to learn on our side. These are whole sentences, typed the way you would say them to a person, and each one is a thing agenthost does.

put this online

Sends the files, works out the runtime, and answers with a real https address. Nothing to set up first.

it needs to save what people type

Adds a database and connects it to the app. You do not see a connection string unless you ask for one.

put a password on it

Locks the door and tells you the password to pass on. Search engines never see the page.

use my own domain

Adds the domain, tells you the one record to paste at your registrar, and handles the certificate.

something's broken, have a look

Reads the running app’s logs, finds the error, fixes the file and deploys again.

There is more of this in the how-to guides, which are written for your agent to read as much as for you.

Whatever it built, it can publish.

You do not have to know what it is made of. Your agent works that out and tells us.

  • framework preset
  • build command
  • output directory
  • install command
  • API keys
  • credit card
  • region

Nothing here can stop and ask you a question.

Every one of these is a setting some other host would demand before it would take your files. None of them exists here, which is why your agent can go from a folder to a working link without checking in.

  • Dockerfiles
  • port settings
  • TLS certificates
  • DNS records
  • CI pipelines
  • health checks
  • scaling rules

The one thing agenthost does expect is an entrypoint per runtime, and the deploy reference is where ChatGPT reads it before the first deploy.

How it connects.

ChatGPT keeps custom MCP servers behind developer mode. Turn that on once, then add agenthost with the address below.

https://agenthost.eu/mcp
1

Turn on developer mode

It lives in ChatGPT's settings, under the section for connectors and plugins. OpenAI has renamed that area more than once, so look for developer mode rather than for a particular menu name.

2

Add agenthost

Add a custom connector, give it the name agenthost, and paste the address above. It speaks streamable HTTP, which is what ChatGPT expects.

3

Sign in with your email

ChatGPT sends you to agenthost, you type the code we email you, and you approve the connection. There is no key or secret to fill in.

Developer mode is available on Plus and Pro. On Business, Enterprise and Edu accounts an admin can allow it, or allow specific connectors. The get started page walks through all of it, with the buttons and the sign-in step.

The usual ones.

Do I need an OpenAI API key, or an agenthost one?

Neither. The connector is added in ChatGPT's own settings, and agenthost signs you in through your browser. There is no key field to fill in and nothing to paste into the chat.

Why is this behind developer mode?

OpenAI puts any MCP server it has not packaged itself behind that switch, because a custom connector can write as well as read. It is a one-time toggle, and it applies to every custom connector rather than to agenthost specifically.

Can ChatGPT update the site later, or only create it?

Both. Ask for a change and the new version replaces the old one at the same address, so a link you have already sent keeps working. It can also put the previous version back, because every deploy keeps the one before it.

What can it do besides deploy?

Create projects, attach a database, set environment variables, add a domain, read the app's logs, roll back, invite people to something private, and open a support ticket if it gets stuck.

The full tool reference →

Is there an API key to paste anywhere?

No. agenthost signs you in through your browser, and your agent never sees a password or a key. There is nothing to copy out of one window and into another, and nothing to rotate later.

What does it cost to try?

Nothing, and there is no card. Your first ten websites are free and stay free. You only pay when the thing you built has to run code on our side, or when you want it on your own domain.

What it costs →

Can I connect more than one agent to the same account?

Yes, as many as you like. They all see the same projects, so you can build something in one and update it from another. Connecting the second one is a single click, because your browser is already signed in.

Where is this actually hosted?

In the EU, by a European company. Your files, your database and your logs stay here, and so does the billing.

Not what you use?

agenthost is a standard remote MCP server, so anything that speaks MCP can drive it. These are the ones with a page of their own.

Point ChatGPT at it.

One address to paste, no card, and the thing it built gets somewhere to live.