Chats: your agents, running on your machine, opened here

Last reviewed: 28 Oct 2026

Chats is where the agents that run on your own computer live: Claude Code, Codex CLI, or a plain shell. They run on your machine, started by the local console the installer put there, and you open them here. Naetive doesn't take a copy of your project. What you type and what the agent prints pass through Naetive on the way, encrypted, and are not stored; see What leaves your machine. ChatGPT and Claude on the web are different: they have nothing to attach to, so they connect from Connect and reach your files through the tunnel instead.

A shell is the exception on this page. It is a plain command line with no agent in it, there because you sometimes want one.

What you need

Two things on the machine, both from the one-click install on Connect: the tunnel, which is how the platform reaches your machine at all, and the console, which runs the agents. If either is missing the page says which, and shows the one command to start it.

You also pair the browser with the machine once, on Connect (Pair this browser) or the first time you open a chat. A login alone cannot open a shell on your computer; the machine decides, by you typing a short code in the Naetive box that opens on it (where no box can show, naetive pair). Once paired, chats from that computer open in this browser without asking again.

A chat runs on the computer that started it. Your understanding and your work are the same everywhere you sign in, and History and Pinned list chats from every computer on the project; to open one, its computer has to be on with its tunnel running.

What the page shows

The top line names the computer: Running on your machine, with how many chats are open, how many are live and how many idle. A search box finds any chat by name, job, commit, folder or computer; press / to jump to it.

Pinned is the chats you starred, up to eight, in your order, the same on every computer. Use a pinned chat's star to move it up or down or to unpin it.

Open now is what is running on the machine right now, live or idle, with the folder each one is in. A chat goes on running when you leave the page.

Start begins a new one: Start Claude Code, Codex CLI or Shell, in the project's folder.

History is every chat on this project, on every computer, newest first, grouped into Today, This week and Earlier. It stays after a chat ends. Filters narrow it to chats Opened in Naetive or From a terminal, or to one computer.

Each row shows the chat's name, its computer, what it worked on, and a word for its state: live, idle or ended. Point at a row and that word becomes the one action: Open for a chat that runs (it goes to that chat, never a second copy) or Resume for one that ended (it picks up where it left off, on its own computer). The arrow at the end opens its details: its job, last action, recent commits, files it touched and threads, plus Rename, Copy resume command, Hide from history and Delete from history. Deleting removes Naetive's record only; the chat itself stays on its computer.

Chats on your other computers

When more than one computer has this project, a line at the top says so and lists them; click one to show its chats. A computer that is off shows as off, with when it was last on. If chats are running on a computer the page is not showing, a line says how many, with Show them.

For a chat on another computer, the page asks that computer whether it is running, and shows running or ended from its answer. Until it answers, the row says checking (or not answering, after a while) and offers Show on that computer instead of Resume, so you never start a second copy.

An open chat

Opening a chat fills the page with it. Each open chat is a tab across the top; Ctrl + Shift + ` starts a new Claude Code tab, and a right-click on + offers Codex or a shell. The list on the left lists what is open now and recent chats, with All chats to go back to the full page and New chat at the bottom.

Each tab has a coloured dot: green when the agent printed something in the last few seconds, grey when it is open and quiet, pulsing while it connects, and amber when the chat ended or failed. Never red.

Beside the chat, a panel says what that agent is doing: its job, the threads waiting on it and what it does next. The button in the tab strip shows and hides it.

Closing a tab and stopping the chat are different things. If the chat is still running, the page asks which you meant: close the tab and the chat goes on, or stop it for good.

Beside Work

With a chat open, going to Work or Understanding puts it in a split beside them, so you can watch what the agent is doing against what the project knows. Ctrl + ` shows and hides it. That pane is a terminal in the ordinary sense, and the word is used there deliberately.

Your own terminal still works

None of this replaces working in your own terminal. A Claude Code window you start yourself connects the same way, appears on Work under Who's here, and commits understanding like any other. Its chats show in History under From a terminal. This page is a convenience, never the only way in.