Works with
Codex in your MacBook notch
Codex is at its best on the long run you walk away from. Noveriq keeps that run visible in the notch, brings the moment it stops for a human to you, and jumps you to the exact terminal tab when you want to be there yourself.
A long run is only unattended until it is not
You give Codex something substantial and go do other work. That is the whole appeal. The catch is that a long run does not finish quietly on a schedule you can plan around: somewhere in the middle it hits a command it needs a human for, or it finishes twenty minutes before you thought it would, and either way it is sitting in a terminal tab behind everything else.
The cost is not the pause. It is the gap between the pause and the moment you happen to look. Noveriq closes that gap: the session's card comes forward in the notch with the exact command on it, wherever you are and whatever is full-screen.
What the run is doing, without opening it
Every running session gets a card: what it is doing right now, how long it has been going, and whether it is waiting on you. The ones waiting sort to the front, so a glance at the notch answers the question you would otherwise have answered by cycling through terminal windows.
Nothing about this is a dashboard you log into. It is on your Mac, in the notch, reading what your agents already write locally.
Answering without breaking your place
When an agent's hook blocks on the local Unix socket waiting for a decision, Noveriq answers over that same socket. The prompt resolves in place: your terminal never comes forward, and nothing is typed into it.
What Noveriq can do for a given session depends on what that agent exposes. Codex is spoken to natively, with full status and decision support where the agent offers the channel for it - but a visible session does not by itself mean every prompt can be answered from the notch. Where there is no decision channel, the fallback is Jump, which puts you in the right terminal and leaves the answering to you.
A destructive shell command is flagged on the card before you press anything, which matters more on a run you have not been watching than on one you have. And there are three ways to say yes: the button allows once, the menu beside it allows for the rest of the session without writing anything to disk, or always, which writes a durable rule into the project's own settings.
Land on the exact tab
When you do want the terminal, Jump takes you to the precise window, tab or split pane that session belongs to - not the application, the tab. It works in Terminal, iTerm2, Ghostty, Kitty, Alacritty, WezTerm, Warp and tmux panes. A terminal with no built-in support can be taught with a single line containing placeholders.
Jump is the one feature that needs macOS Accessibility permission, because focusing another application's window is what that permission governs. Answering from the notch does not need it.
Setup is one launch
Hooks are configured the first time Noveriq runs, and removed again when you delete the app. There is no file for you to edit and nothing left behind afterwards.
# nothing to run. Open Noveriq once, then start Codex
# the way you always do:
codexThe next time that run stops for you, the card is in the notch.
It stays on your Mac
Sessions are read from what the agents already write locally, and decisions travel over a local socket. There are no accounts, nothing to log into, and no analytics in the app at all. Transcripts, prompts and diffs never leave the machine, and no session is visible from another one.
What it does not do
Noveriq shows how much of your subscription window is left, read from the Codex CLI's own rate limit reporting, plus how much context each session has left. What it does not do is count tokens or report cost: if you want a breakdown of what a run has spent, that is a different question and a different tool. It is not a terminal, not an IDE, and not a way to run agents - it watches and answers the ones you started yourself.
Questions
Does it handle several Codex sessions at once?
Yes. Each running session gets its own card, and the ones waiting on a human sort to the front.
Does answering from the notch type into my terminal?
No. The answer travels back down the socket the hook is already blocking on. There is deliberately no synthetic-keystroke path anywhere in the app.
Does it tell me what a run cost?
No. It shows how much of your session and weekly windows is left, from the Codex CLI's own rate limit reporting, but it never counts tokens or reports a dollar figure. For a spending breakdown, ccusage is the tool for that.
Does it work with other agents at the same time?
Yes. Claude Code, Cursor, Gemini CLI, Qwen Code and opencode are all spoken natively, and any other agent CLI is detected while it runs and gets a card you can jump from.
Walk away from the run, not from the decision
One price, one Mac, no subscription. Download it first - it asks for a key when you are ready.
Get Noveriq for Mac$9.99 once · macOS 15 or later · 30 days, money back