What Guard collects
Guard exists to find accidentally exposed services. It is deliberately small: it does not watch what you do.
What is collected
- Your account details and the contact information you enter in your profile.
- For each connected device: its name, hostname, operating system, client version and when it last reported.
- The list of network services the client says it is listening on: port, protocol, bind address and the process name.
- The public address your connection came from, each time you sign in or the client reports.
- A short log of security-relevant events: sign-ins, device connections, revocations and internet checks.
What is not collected
- Files, documents, browser history or installed software.
- Screen contents, keystrokes or microphone and camera.
- Network traffic or anything about other devices on your network.
- Your location, other than the network address your connection came from.
How the internet check works
When a client reports, Guard connects to the address your request came from — never to an address the client names — and tries only the TCP ports that client declared. It opens a connection and closes it again; it sends and reads no data.
Guard checks what a client declares. A modified or stopped client is not seen, so a device that stops reporting is flagged instead of being shown as secure.
Who can see it
You can see everything about your own devices and findings. Administrators in the security team can see each employee's status and the event log; opening that view is itself logged.
How long it is kept
Events and resolved findings are deleted after 180 days.