Sharing what an AI tool builds — Claude artifacts, vibe-coded apps, Codex bundles — before you deploy.
It looked like a URL. You sent it. It doesn't exist — and never did. Why models produce link-shaped text when you ask them to share something, how to spot it in one second, and what actually has to happen for a real link to exist.
The agent handed you a URL and said it's live. It probably is. What it didn't tell you is who can open it, how long it lasts, whether the address is guessable, and what happens to it when you revise — four defaults chosen for you, none of them announced.
ChatGPT Sites hosts what you build and gives you access controls — owner only, named workspace members, the whole workspace, or the public internet. There's no rung for one outside client. Here's the shape of that gap and the route round it.
The steps take about ten seconds. The cost is less obvious: unpublishing permanently retires that artifact's ability to be published, and deletes its stored data. Here's how to do it, what it takes with it, and what to reach for when access only needs to end for now.
The normal delivery ritual is finish, then send — and every revision mints a new URL and another 'ignore my last email'. If your AI tool can publish and then update in place, the link can go out first and stay correct. Including when that's a bad idea.
Claude refuses to share your artifact and suggests publishing a new version — which often doesn't clear it. Here's the diagnosis that actually separates the two failures behind that message, and the export route that doesn't depend on the toggle working.
Not to a public link — on any plan. On Pro and Max, a connector-backed artifact can't be shared at all. Here's the mechanism that makes that rule inevitable, and what to do once you know which page you're actually building.
The iteration loop with Claude / Cursor / Lovable produces a new version every round. Channels that let each round travel cleanly — without exposing the chat, the previous version, or the next reviewer to noise.
A step-by-step: turn on Review for a private link, let anyone pin comments to the exact spot (no account), then export the thread as a prompt to paste into Claude or Codex.
You shared the AI-built page — then what? The real work is the loop: build, show, react, fold the reaction back into the next prompt. Why hosting is only step one, and what actually closes the loop.
"It asks me to log in." "It's blank." The link isn't broken — it was never built to leave the tool. What each AI tool actually hands you, and what opens.
An AI tool builds the landing in fifteen minutes; deciding where it lives takes the next hour. Three hosting shapes and how to pick between them.
Sending a Claude or ChatGPT artifact to one client — how the channels compare, where private-link hosting fits, and the six clicks that get you there.