# My AI says it built it, but I cannot see anything. What now?

Checked 2026-11-15 against GilbApp d5296208bbffa2ab70076e6194f5537c4be1ddf7.

Ask your AI to start your app on your computer and give you the address; a live app changes only after your go.

## Deeper

Built is not the same as live, open to the people who will use it. Your app's code is kept in a folder on your computer, and your AI writes it there. Kai-Zen, the planning agent, puts a step live only with your go. Until then, your app's address on the internet, https://&lt;name\>.pages.dev, shows the version before, or nothing if it was never put live. The answer [How can I see what I've built?](http://support.gilb.com/help/how-can-i-see-what-ive-built) says more.

First checks.
Is your app running? Ask your AI to start your app on your computer and tell you the address to open in your browser. Such an address works only on your computer, while the app runs there; no one else can open it. If it does not open, or shows the old version, ask your AI to start it again, then reload the page in your browser.
Did your AI change any files? In GilbApp's chat, with [Claude Code connected](http://support.gilb.com/help/how-do-i-connect-my-ai-to-gilbapp-chat) and "Code in" switched on, the line under your AI's answer says "Changed in" and your folder's name, with how many files were added, changed or removed. "No files changed in" and the folder's name means that answer changed no file. Grok in the chat does not build in a folder on your computer: the code folder is for Claude Code only. In a terminal, the window where you type commands, ask your AI which files it changed. The answer [How do I see what changed, and who changed it?](http://support.gilb.com/help/how-do-i-see-what-changed-and-who-changed-it) shows how.
Are you signed in? An app on GilbPlatform reads and writes only what the signed-in person may reach, so the same app can show nothing for one person and work for another. A part made for signed-in people shows nothing until you sign in to your app.
Is it data, not a screen? What your app saves is kept as records in a vault, not in the app. The answer [Where does my app keep its data, and can I see it?](http://support.gilb.com/help/where-does-my-app-keep-its-data-and-can-i-see-it) shows how to look.

[Read the handback](http://support.gilb.com/help/how-do-i-read-what-an-agent-hands-back), the message your AI ends a step with. When a step changes a screen, Agent.Coder, the agent that writes your app's code, must also open that screen in a real browser at phone size, take a screenshot and count the errors before it hands the step back. If the screen is blank or the browser reports errors, it must fix it and check again. Its handback gives where that picture is saved, so you can open it. A passing test run is not this proof. If a step changed a screen and its handback has no picture, ask Kai-Zen whether the screen was opened in a browser; if the handback says there was no browser tool, the agent that handed the step over, Kai-Zen, takes that check. An agent is your AI following one written procedure, its standing instructions, for one kind of work.

When the plan is kept in a vault, the step's card in GilbApp's "Plan" shows its "Status", where it stands: "Building" while it is being built, "Developed" only after the testing agent has passed it, and "Production" only once it is live.

Then tell your AI. Say what you opened, what you expected to see and what you saw, with the exact words of any message on the screen. For example: You said the order list is built, but the page at the address you gave me is empty. Agent.Coder reads the browser's error messages itself with its browser tools; its instructions say not to ask you to copy them from the browser, unless no tool works. The answer [How do I tell my AI about a fault so it can fix it?](http://support.gilb.com/help/how-do-i-tell-my-ai-about-a-fault-so-it-can-fix-it) shows what a good report says.

This answer is written for a computer screen.

Source: Standards: Agent.Coder.AfterIncrement \(a real browser at phone size when the step changes the UI, screenshot, console error count, "A passing test run is not this proof", the screenshot's path in the return, no browser tool said\), Agent.Coder.Browser \("Do not ask the requester to copy from the browser"\), Kai-Zen.Cycle.Check \(Developed only after the tester's saved PASS\), Kai-Zen.Cycle.Live \(Production only after the release is proved live\), Kai-Zen.GoLive.Ask and Kai-Zen.GoLive.Deploy \(https://&lt;name\>.pages.dev, only after the person says yes\). GilbApp 49b1ae53 \(https://gilb.com/app/version.json, read 13 November 2026\): src/components/agent/CodeFolderParts.tsx:57 \("Changed in", "No files changed in"\); src/components/agent/AgentChat.tsx:393 \("Code in"\), :677 \(code in a folder only with Claude Code\); src/lib/base/planCard.ts:31 \("Building", "Developed", "Production"\). Live answers, sentences shared: How can I see what I've built?; How do I see what changed, and who changed it?; Does my computer have to stay on while my AI builds? \(Grok and the code folder\); Where does my app keep its data, and can I see it?; My AI says the tests passed, but it does not work for me. What now? \(the signed-in person\); How do I tell my AI about a fault so it can fix it?.
