Work · Community report
Set up SSH access for your agent across computers
Set up SSH access for your agent across computers. Give the agent permanent SSH access to your Mac and other computers over a Tailscale network.
What you can get
What it is
SSH connection setup
Standing SSH access from the agent to the computers on your Tailscale network, so it can run commands on them without you setting up a connection each time.
It came back right when
- SSH connects successfully to each intended computer over Tailscale.
- Authentication method matches what you decided on, not a weaker fallback.
- You know which commands can run unattended and which need your approval.
The checks are editorial, not from the source.
- Confirm Tailscale is workingStops for your approval
Verify network connectivity across target computers first.
- Choose authentication methodStops for your approval
Key-based or password, decided deliberately.
- Set up and test SSH access
Connection verified on each individual machine.
- Set command limitsStops for your approval
You decide what runs unattended versus with approval.
Editorial, not from the source
Works when
- You already run Tailscale across your own computers.
- You're comfortable setting up and securing SSH access yourself, or reviewing how it's configured.
- You want the agent to reach your machines without reconnecting each session.
Does not work when
- You don't yet have Tailscale or SSH set up and want this as a first introduction to either.
- You're not comfortable with an agent having standing access to run commands on your computers.
- You share these computers with others who haven't agreed to this access.
Agent SSH access: review template
| Item | What it is based on | Status |
|---|---|---|
| Tailscale connectivity | Verified across target computers | [Verify] |
| Authentication method | Your deliberate choice, key or password | [Confirm] |
| Command limits | What runs unattended vs. needs approval | [Review] |
The source case
On X, @JadenKwek reports setting up permanent SSH for their Muse agent to their Mac and all their other computers over their Tailscale network. This is one attributed report, checked on 2026-10-09: it does not describe the specific setup steps, whether key-based or password authentication was used, how many computers were connected, or what commands the agent is actually used to run once connected.
A public report by someone outside this library. Not independently reproduced here.
From the source
Source excerpt
I set up permanent SSH for my muse agent to my Mac and all my other computers on my tailscale
Quoted from the original posts. Not written by this guide, and not a result produced here.
Suggested prompt
Bring A working Tailscale network across your computers, SSH enabled on each target computer. Acts Connects to your computers over SSH via your Tailscale network. Runs commands on connected machines when you ask it to. Source community report, checked 2026-10-09.
Written by this guide as a starting point, not quoted from the source. Your edits stay in this browser.
Other ways to ask
Editorial rewrites of the brief above, for a different emphasis. Copy them separately.
What you need
Editorial, not from the source
You supply
- A working Tailscale network across your computers
- SSH enabled on each target computer
Connected
- Tailscale network access
- SSH credentials or keys for each computer
Approved by you
- Your decision on which commands run unattended versus with approval
- Your review of which computers have standing access
How to try it
Editorial, not from the source
- Confirm Tailscale is already set up and working across the computers you want reachable.
- Enable SSH on each of those computers, and decide whether to use key-based or password authentication.
- Have it set up a permanent SSH connection to each machine over the Tailscale network.
- Test the connection on each computer individually before relying on it for real tasks.
- Agree on what kinds of commands it can run unattended versus what needs your approval first.
Start from a working Tailscale network
This task assumes Tailscale is already connecting your computers to each other; it's not a guide to setting up Tailscale itself. The attributed report describes permanent SSH being set up for the agent across a Mac and other computers already on the author's Tailscale network, but gives no detail about how that network was configured in the first place. If you haven't already got Tailscale running across the machines you want the agent to reach, get that working and verified first. Trying to layer SSH access onto a network connection that isn't solid yet just makes any problem harder to diagnose later, since you won't know whether a failure is in Tailscale or in SSH itself.
Decide on authentication deliberately
The attributed report doesn't say whether key-based or password authentication was used for the SSH connections it set up, which means there's no established pattern to copy here. Key-based authentication is generally considered more secure for a connection meant to be permanent, since it avoids a password that could be guessed or reused elsewhere. Whichever you choose, make the decision deliberately rather than defaulting to whatever is easiest to set up quickly. Standing SSH access to multiple personal computers is worth a few extra minutes of setup, especially since this access is meant to persist rather than being created and torn down for a single task. Write down which method you chose and why, since that note is what lets you review the decision later instead of guessing at it again.
Decide what it's actually allowed to run
Having SSH access and having unattended permission to run anything are two different decisions, and the attributed report doesn't describe what commands the agent actually runs once connected. Before relying on this setup for real tasks, work out which kinds of commands are fine to run without asking, read-only checks or routine status queries, for example, and which ones should require your approval first, like anything that installs software, deletes files or changes system settings. Writing this down for yourself, even informally, makes it much easier to notice later if the agent's behavior starts to drift from what you originally intended, and gives you something concrete to point back to if a change needs explaining.
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
- One attributed report, and a short one. It does not describe the actual setup steps, authentication method, or number of computers connected.
- It does not say what commands the agent is used to run once SSH access exists, or whether access is limited to specific directories or actions.
- Permanent remote access of this kind to multiple computers is a real security surface; this directory has not reviewed how access was secured in the attributed case.
- Tailscale and SSH configuration details vary by operating system and setup; expect to work through platform-specific steps not covered here.
Stop and check
Editorial, not from the source
- Before setup: confirm Tailscale and SSH are already working on each computer.
- During setup: decide on authentication method and which commands can run unattended.
- After setup: test the connection on each machine and periodically review which computers still have standing access.
If it goes wrong
Editorial, not from the source
- SSH connects to some computers but not others
- Check that Tailscale and SSH are both enabled and reachable on each machine individually; the attributed report doesn't detail a troubleshooting process.
- You're unsure how secure the setup is
- Review the authentication method yourself, key-based is generally stronger than password-based, since the source report doesn't specify which was used.
- The agent runs a command you didn't expect
- Tighten what's allowed to run unattended and require approval for anything beyond routine checks.
Common questions
Do I need Tailscale specifically for this?
The attributed report describes the setup over a Tailscale network. A different mesh VPN or network approach could work similarly, but this page is based only on the Tailscale case reported.
Is key-based or password authentication used?
The source report doesn't say. Decide this yourself; key-based authentication is generally considered the stronger choice for a connection meant to be permanent.
What commands does the agent actually run over SSH?
Not specified in the attributed report. Decide in advance what's fine to run unattended and what needs your approval first.
How many computers were connected in the report?
Not stated. The author mentions a Mac and other computers on their Tailscale network, without a specific count.
Is this secure?
This directory has not reviewed the security of the attributed setup. Standing remote access across multiple computers is worth securing deliberately, with authentication and command limits you decide on yourself.
Next
Related to this task
More in Work
Get recurring task recommendations — ask your own agent to check for relevant updates.