Built for people working in parallel

Your AI agents are waiting on you.
Now you know when.

Signal Lights shows every Codex and Claude Code task in your Mac menu bar, so you can see what’s working, what needs you, and what’s ready without hunting through windows.

Native macOS app · Apple silicon · macOS 13 or later

Signal Lights by JupiterAI

1 task needs you
Live
AReview launch checklistWebsite launch · Needs your attention
Refine onboarding flowSignal Lights · Working
AAudit release buildMac app · Finished · unread
Add launch at loginSignal Lights · Finished

Waiting for youReview launch checklist

Ready to readAudit release build

One calm view across OpenAI CodexA Claude CodeRuns locally on your Mac
The problem

The hardest part of parallel work is remembering what’s waiting.

You start one task, jump to another, answer a message, then discover an agent has been waiting for approval for twenty minutes. Signal Lights keeps the state of your work visible, even when the windows aren’t.

Four states. Zero guesswork.

Glance once. Know what to do next.

01

Needs your attention

An approval, question, or decision is holding up the work. Red stays visible until you respond.

02

Working

Your agent is moving. A warm pulse gives you confidence without demanding attention.

03

Finished, unread

The result is ready. Blue remains until you open the task, so completed work never disappears.

04

Finished and read

You’ve seen the result. The task settles into green and gets out of your way.

Designed for flow

Stay in the work you’re doing.

Signal Lights quietly watches local task state and brings only the useful part into view. Click any task to return to its conversation.

  • Connect your assistantChoose Codex, Claude Code, or both during onboarding.
  • Keep movingYour menu bar updates as tasks work, wait, and finish.
  • Act at the right momentUse colour and optional sounds to spot the next useful action.
121Everything in sight.Nothing shouting.
Codex
A Claude Code
Your menu bar

Local by design

Your conversations stay on your Mac.

Signal Lights reads bounded local status metadata. It does not upload your prompts, tool inputs, tool results, or conversation bodies. Optional anonymous usage sharing is off by default. No monitoring account. No cloud copy of your work. Read the Privacy Policy.

Stop losing the thread

Give every AI task a place in your peripheral vision.

Install once, connect the tools you use, and let Signal Lights keep watch while you focus.

Signal Lights for MacDirect download

A$49.99 / year

  • Codex and Claude Code integration
  • Live task states and unread tracking
  • Clickable task links and sound alerts
  • Launch automatically at login
A$49.99 charged annually. Renews automatically until cancelled. Cancel before renewal; access continues to the end of the paid period. Secure checkout powered by Stripe.By subscribing, you agree to the Terms & Conditions and acknowledge the Privacy Policy.

Questions, answered

Before you switch it on.

Does Signal Lights read my conversations?

No. It reads bounded local status metadata needed to determine whether a task is working, waiting, finished, or read. Conversation bodies are not copied or displayed.

Can I use only Codex or only Claude Code?

Yes. Onboarding connects the assistant you use. You can add the other integration later if your workflow changes.

Does it work with Claude Code in Terminal?

Yes. Official Claude Code hooks report local lifecycle events from terminal sessions. Some terminal-only sessions cannot deep-link back into Claude because Claude does not expose a link for every CLI session.

Can I use Signal Lights offline?

Yes. A successful subscription check keeps the app available offline for seven days. Signal Lights checks at most once a day and resumes automatically after you reconnect.

How do I cancel or change my plan?

Choose Manage Subscription in the Mac app to update your card, see invoices, cancel at the end of your billing period, or schedule a monthly or annual plan change for renewal.

What Mac do I need?

The current release supports Apple silicon and Intel Macs running macOS 13 or later.