Work · Community report
Unlock your Mac remotely with a message
A Mac owner on X reports building a shortcut that unlocks the machine while they are away once their Muse agent messages "unlock", and passing terminal jobs to Codex. The post shows no setup details or result, so this page lays out how to unlock your Mac remotely with a message without opening a security hole.
What you can get
- Prepare
A Mac you own and administer, with its macOS version
- Ask and review
Decide whether a remote unlock is worth the risk for this Mac; a work machine may be covered by your employer's security policy.
- Check the result
A written setup plan and safety checklist for a message-triggered unlock shortcut; not a verified working automation or a security review of your Mac.
The source case
@scootfleabag reports setting up a shortcut that unlocks their Mac while they are away whenever their Muse agent sends the message "unlock". They also say they taught the agent that, when it needs the terminal, it should send a message to Codex and have it do the work. The post does not say which app or script runs the shortcut, how the sender is checked, or whether the unlock has been used successfully; it shows no screenshots or setup steps.
A public report by someone outside this library. Not independently reproduced here.
From the source
Source excerpt
Set up a shortcut that unlocks my Mac when I’m away if my muse agent sends the message “unlock” Also taught it if it needs to use the terminal, just send a message to codex and have it do what it needs
Quoted from the original posts. Not written by this guide, and not a result produced here.
Suggested prompt
Bring A Mac you own and administer, with its macOS version, The messaging app and exact trigger word you want to use, A decision on who may send the trigger and how to disable it. Source community report, checked 2026-10-22.
Written by this guide as a starting point, not quoted from the source. Your edits stay in this browser.
What you need
You supply
- A Mac you own and administer, with its macOS version
- The messaging app and exact trigger word you want to use
- A decision on who may send the trigger and how to disable it
How to try it
Editorial, not from the source
- Decide whether a remote unlock is worth the risk for this Mac; a work machine may be covered by your employer's security policy.
- Pick the trigger: one messaging app, one exact word, and only messages from the agent's own account or number.
- Have the agent write out the setup steps and the permissions each tool asks for, then do each step yourself.
- Keep your login password out of plain-text notes and chat; if a tool needs it, use the macOS keychain or skip that tool.
- Add an off switch you can reach from your phone, then test once while you are sitting at the Mac.
- If you let the agent pass terminal work to another coding tool, require it to show each command and wait for your approval.
What the post describes
@scootfleabag writes that they set up a shortcut on their Mac which unlocks it when they are away, triggered when their Muse agent sends the message "unlock". In the same post they say they taught the agent to send a message to Codex whenever it needs the terminal, and add that AI using AI as a tool feels like a glimpse into the future.
That is the whole account. It does not say which automation tool listens for the message, how the password is supplied, or whether the unlock has worked in practice; treat it as a reported setup rather than a recipe.
Security questions to settle first
In general, macOS is designed so that unlocking needs the user's password, Touch ID or a paired Apple Watch, and remote unlock tricks usually work by storing the password somewhere a script can reach it. Before copying the idea, ask who else can send a message to that inbox, what happens if your phone or the agent's account is compromised, and whether the Mac holds work data covered by a company policy.
A narrower setup is often enough: a remote screen-sharing session you approve each time, or waking the Mac without unlocking it.
Open Muse to try it. That is the product's own site; this guide is not part of it.
Limits and confirmation points
Limits to confirm
- The post does not name the app, script or settings behind the shortcut.
- No result is shown: there is no screenshot, log or report of the Mac actually unlocking from a message.
- The hand-off to Codex is described in one sentence; the post does not show what commands it ran or how it was asked.
- Anyone who can send the trigger message may be able to unlock the Mac, so the setup carries a security risk the post does not discuss.
- This is an attributed community report, not reproduced by this guide. The suggested workflow and checks are editorial guidance, not a recovered conversation or guaranteed product capability.
If it goes wrong
Editorial, not from the source
- The shortcut runs when someone other than the agent sends the trigger word
- Turn the automation off, then add a check on the sender's account or number before re-enabling it, or change the trigger to something less guessable.
- The Mac wakes but stays on the login screen
- Many automation tools cannot type into the macOS login window without accessibility permissions; check System Settings > Privacy & Security, and decide whether granting them is acceptable.
Common questions
Can Muse unlock my Mac?
One user reports that a shortcut on their Mac unlocks it when their Muse agent sends the message "unlock". The post does not show how the shortcut is built or that it has run, and this guide has not reproduced it.
Is a message-triggered unlock safe?
It depends on who can send the trigger and where the password is stored. Limit the trigger to one sender, keep the password out of plain text, and keep a quick way to turn it off.
What does handing terminal work to Codex mean here?
The author says they taught the agent to message Codex when it needs the terminal and have Codex do the work. No commands or results are shown, so approve each command yourself if you set up something similar.
Next
Related to this task
More in Work
Get recurring task recommendations — ask your own agent to check for relevant updates.