KairosLink
Getting started
From a brand new account to your first device reporting: create your first customer, install the agent, and get monitoring alerts actually reaching you.
By the end of this guide you will have your first customer created, one of their devices reporting in the console, and alerts set up so they actually reach you.
It takes about fifteen minutes.
Your account
When you sign up you start with a 14 day trial, no credit card. During that time you get every module enabled, up to 10 devices and 3 technicians.
The first thing you see is not the console but a screen asking you to confirm your email. Find the message and click the link. If it never arrived, that same screen has Resend verification email.
When the 14 days are up the service is not cut off. The account moves to a paid plan, so if you are staying, add your payment method under Subscription.
The order
This is what trips people up most at the start, so here it is plainly:
Customer, then installer, then device.
You do not add devices by hand. Devices appear on their own once the agent is installed and connects. And the agent is installed with an installer generated from a customer, so the device is assigned to that customer from the first second.
If you are an MSP with twenty customers, each one is a customer here. If you are the IT department of a single company, create one customer with the company name and you are done.
Step 1: create your first customer
- Side menu, Customers.
- Click + New customer.
- Type the Customer name. It is the only truly required field: status already comes set to Active.
- Create customer.
Everything else (contact person, email, phone, address, notes) can be filled in later. Do not stall here.
Two fields worth understanding but safe to leave for later:
- Customer language affects the portal and the emails that customer receives. It does not change your console language.
- Monthly hour bank only matters if you are going to use time tracking. It resets on the 1st of each month.
Step 2: install the agent
Open the customer you just created and go to the Computers tab. The Installer section is down the page.
There are six paths. Pick by your situation:
| Your situation | Path |
|---|---|
| Windows, you are at the machine or can reach it | Download the MSI |
| Windows, and you need someone at the customer to install it | Copy the public link |
| Windows, many machines and there is a domain | GPO deployment |
| Linux with internet | The one-line command |
| Linux with no internet access | The four files |
| macOS | The one-line command |
For your first device, go with the MSI or one of the two one-line commands.
Windows: the MSI
- Click + Create Installer.
- Optionally fill in the site, department and device type tags. The device will be classified from the start.
- Wait. Generation takes between 30 and 90 seconds: the installer is compiled with your customer's credential inside it.
- When it is ready, download it and run it on the machine.
The installer is valid for 7 days and works on more than one machine during that time. You do not need one per device.
If you are not at the machine, use Copy link. That link downloads the installer without asking for a login, so you can send it to someone at the customer by email or chat without giving them access to your console. It expires with the installer.
Linux: the one-line command
- In the dialog, switch to the Linux tab.
- Click Copy command.
- Paste it on the server and run it.
It is a single line that installs and enrolls in one go. Nothing to compile or download beforehand.
Requirements: the machine must use systemd and be x86_64. The installer covers Ubuntu, Debian, RHEL, Fedora, CentOS, Rocky, AlmaLinux, Oracle Linux, SUSE, openSUSE, Alpine, Amazon Linux and Arch. Alpine only works if it is set up with systemd, since it uses OpenRC by default.
If you run the same line again on a machine that already has the agent, it updates and does not duplicate the device in the console.
macOS: the one-line command
- In the dialog, switch to the macOS tab.
- Click Copy command.
- Paste it on the Mac and run it.
As on Linux, it is a single line that installs and enrolls in one go, and it does not expire. The binary is universal: the same download covers Apple Silicon and Intel.
Remote control needs one more step on the Mac: the user of the device has to grant three system permissions, and they cannot be granted remotely. It is covered in the Install the macOS agent guide.
Windows at scale: GPO
If the customer has a domain and you already have the agent on the Domain Controller, you can deploy to everything else at once, touching no machine and using no credentials.
You need four things: the Active Directory module (from the Business plan), the agent deployment permission, a Domain Controller that is online and assigned to a customer, and an MSI already built for that customer.
Installation happens at the next policy refresh, between 90 minutes and 2 hours, or sooner if the machine reboots. Up to 500 devices per campaign.
This is the only path with its own diagnostics screen: you see how many enrolled out of the total and what each one reported.
Step 3: see the device in the console
The device appears online the moment the agent connects. There is no cycle to wait for.
What does arrive in stages is the data:
| Data | When it shows up |
|---|---|
| Name, operating system, CPU, RAM, IP | Immediately |
| CPU, RAM and disk metrics | Within 5 minutes |
| Detailed OS and installed software | Within 30 minutes |
| Pending patches | On the nightly scan at 1:00 AM |
| Certificates | On the daily pass at 2:30 AM |
So if ten minutes in the device is online but has no software list, everything is fine. It is on its way.
A device is marked offline after 70 seconds with no signal, and the check runs every minute. In practice, between 70 seconds and two minutes.
The first thing to do afterwards
Turn on email alerts for the device.
This is the step people skip most and the one that costs most. A freshly installed device is monitored from minute zero: it raises alerts in the console, records metrics, detects outages. But it does not send a single email until you flip that switch, device by device.
If you skip it, monitoring works and you find out about nothing until you open the console to look.
The switch is in the device record, in the monitoring section.
After that, the order that makes sense:
- Thresholds and alerts, under Settings, Monitoring section. You set the CPU, RAM and disk percentages that should trigger a warning, and how often checks run.
- Branding, under Settings. Your logo and colours on reports, emails and the customer portal. This is what makes your customer see your brand instead of ours.
- Users and roles, to add your team. Remember the trial allows up to 3 technicians.
- Customer portal, when you want your customer to see their device status and open tickets themselves.
- Customer notifications, to set which addresses receive each customer's alerts.
I installed the agent and the device does not appear
Check in this order:
Are you looking at the right customer? The device shows up under the customer whose installer you used, not anywhere else.
Have you hit the device limit? The trial allows 10. The Managed devices card on the dashboard shows your count. At the limit, enrollment is rejected and the device never appears.
Has the installer expired? It lasts 7 days. After that you need to generate a new one.
Did you rotate the customer's credential after installing? If you rotated the enrollment credential, agents installed with the previous one can no longer report.
If it came through GPO, open the campaign detail: it lists which machines are missing and what each one reported.
If it was Linux or macOS, the installer prints the reason for the failure in the same console where you pasted the command. Go back and read that output.
If the device appears but went offline, enrollment worked and it is the keep-alive signals that stopped. Check that the agent service is running on the machine.
Frequently asked questions
Can I add a device manually? No. Devices appear on their own when the agent connects.
Do I need one installer per device? No. One covers every device for that customer for 7 days.
Does an installer work for another customer? No. Each installer carries its customer's credential and enrolls against that customer.
Can I move a device from one customer to another? Yes, from the device record. Bear in mind it breaks the link to the original credential, so reinstalling later with the old installer will create a duplicate.
What if I run the installer twice on the same machine? On Linux and on macOS the agent updates and does not duplicate. On Windows, uninstall first.
Do I need to open firewall ports at the customer? No. The agent reaches out; it does not accept incoming connections.
How long is the trial and what does it include? 14 days with every module, up to 10 devices and 3 technicians, no card.
What happens on day 15? The service is not cut off. The account moves to a paid plan. Add your payment method under Subscription if you are staying.
Why is monitoring not emailing me? Because email alerts start switched off on every device. You have to turn them on.
Can I install the agent on a machine with no internet? On Linux yes, using the direct download of the four files, and on macOS using the direct download of the agent files. On Windows the machine needs to reach the console to enroll.
Updated August 21, 2026