Skip to content
How device management works

Profiles, policies, monitors and software, in full.

Configuration profiles that put a setting back when it drifts, compliance policies that say what happens when a device falls short, monitors that fix what breaks and BitLocker keys you can find when you need them. For every client, from one section, beside Intune rather than against it.

secuno / devices / profiles
Workstation baselineSTAGED ROLLOUT
Pilot devices
Passed
10% of the rest
Passed
50% of the rest
Soaking
100%
Waiting
Halts when
Over 5% fail or go silent

Each stage soaks before the next one starts.

0
Settings in the catalogue, for Windows and Linux
0min
Minutes between checks for drift on every device
0%
Percent of failing devices that halts a rollout
0h
Hours after a password is revealed, it is rotated
One section, every client

The fleet at a glance, client by client.

The Devices section covers every client you can reach, with a client filter that follows you through it. The overview has one row per client, worst first, so the question "where do I start today" has an answer.

What each row tells you

Compliance with your profiles, reporting devices and agent coverage, failing and conflicting settings, open and critical alerts, app installs that failed or wait, and how many devices have their keys escrowed. Totals and the worst settings across clients sit above it.

Aim once, reach the right devices

Profiles, compliance policies, monitors and software target chosen clients, every client of a partner (including clients added later) or the whole platform. Inside a client you include all devices, device groups, tags or named devices and exclude others, and an exclusion always wins. A reusable assignment filter narrows it further by what a device is: OS version, architecture, join type, manufacturer or model.

Built for teams that share clients

Someone who reaches only some of a rule's clients sees it read-only, with the other clients counted but not named. Partners get a Standards matrix of clients by partner-wide profiles, showing where a client overrides a setting and where devices drift.

Configuration profiles

Settings that stay set.

A profile is a set of settings from a fixed catalogue. The agent applies them, checks them every 15 minutes and puts them back when something changed them. You see per device whether each setting holds, and which profile it came from.

  • Pick a type when you create a profile: settings catalogue, an endpoint security template or a security baseline for Windows workstations, Windows servers or Linux servers. When a baseline changes, its profiles are offered the new version for review
  • Before you save, an impact preview shows per setting on how many devices the profile would win, lose to another profile, change the value or be blocked by Group Policy or an MDM
  • Report mode only checks: use it before you enforce anything that could lock people out
  • Where two profiles set the same setting, the lowest priority number wins, and a group profile beats a client-wide one on a tie
  • A change reaches devices when you save it or in each device's maintenance window, your choice per profile. Settings that need a restart read pending until then and can use the same restart plan as patches
  • Take a setting out of a profile and the agent puts back the value from before Secuno changed it. You can also choose to leave devices as they are
  • Every state change of a setting on a device is kept for 180 days as a history timeline
  • A device where an enforced setting fails gets one risk, resolved on its own once nothing fails
Windows17
FirewallDefender protection19 ASR rulesBitLocker + escrowManaged local adminLocal AdministratorsScreen lockUSB storageSMB signing, SMBv1LLMNR, NetBIOS, WDigestLSA protectionRemote Desktop, NLAUAC, AutoRunAudit policyScript block loggingPassword + lockoutRegistry values
Linux09
USB storageSSH root + passwordsSSH hardeningKernel + network sysctlBlocked modulesDisabled servicesPassword policyCore dumpsauditd
Checked, never written03
Secure BootTamper ProtectionLinux host firewall
Custom registry values are allowed, except protected keys. The Linux host firewall passes only when traffic is actually dropped.
Plays well with others

Secuno never fights another tool.

Most clients already have a GPO, an MDM or a previous admin's script. A management tool that overwrites them every 15 minutes causes more incidents than it prevents, so Secuno steps back instead.

Managed elsewhere

A setting a Group Policy or MDM policy owns reads managed elsewhere and is not written. This is judged per ASR rule, per sysctl file and per SSH drop-in, not for a whole area at once.

Beside Intune, not against it

Secuno reads each connected tenant's Intune configuration profiles, settings catalogue, endpoint security and compliance policies every six hours, and never changes them. You see each Intune policy beside the Secuno settings it touches, and the profile preview warns when Intune already sets a setting on the clients you target.

Conflict, not a tug of war

A setting that something changes back three times in a day reads conflict and is left alone until the profile changes. Routine work such as signature updates and scheduled rotation never counts as drift.

Guard rails on the device

A default inbound block never loosens a profile that already blocks outbound. Turning SMBv1 off keeps the SMB client starting after the restart. Accounts read not applicable on a domain controller, and the agent refuses to disable drivers, system services or anything it depends on itself.

Leaves the device as it found it

The agent records a value before its first change and restores exactly that when the setting leaves the device. For lists such as ASR rules, services and local administrators it records each item, so items added later go back too. Every setting states what a revert restores: full, partial or none, with the reason.

Why this value?

Every setting on the device page answers it: each profile that sets it in order of precedence, why the winner won, the compliance policies that check it and its last 20 history events.

Staged rollout

A bad setting meets ten devices, not ten thousand.

Every change to a profile is a new version that moves through stages, each held for a soak time. A gate checks every five minutes and stops the rollout the moment the change starts doing harm.

How a profile version reaches the fleet
PilotDevices tagged pilotYour own canaries, chosen by tag
10%A tenth of the remaining devicesHeld for the soak time before moving on
50%Half of the remaining devicesStage sizes are configurable
100%Every targeted deviceOnly after every earlier stage held
HaltMore than 5% of reporting devices fail a changed setting or go silent right after receiving itThe rollout stops at once and the profile's owners are told

Pause, resume, advance or publish now when you are sure. Abort sends the published values again as a new version. The console shows each stage, the gate's numbers and every version with what changed. New profiles are staged by default. Software assignments and monitors move through the same stages: software halts on failed installs or devices that go silent, a monitor when too many devices in a stage breach or error.

Compliance policies

State the rule. Decide what happens when a device breaks it.

A profile enforces. A compliance policy judges: it says what a device must meet and what follows when it does not, and it changes nothing on the device. Devices are evaluated every 15 minutes, after each save and on request.

Rules on what you already measure

Any device compliance check, such as BitLocker, Secure Boot, TPM, firewall, antivirus and signatures, UAC, patches current or an operating system still in support. A catalogue setting holding a given value, checked without enforcing it. A minimum or maximum OS version. The agent checked in within a number of days.

Grace first, consequences after

A device is compliant when every rule passes, in grace when one fails, then noncompliant. You schedule what follows in days from the first failure: mark it noncompliant, notify the client's admins with the devices named, raise a risk per device that resolves when it passes again. A rule with no data reads not evaluated, never compliant.

Block noncompliant devices in Entra

Per Microsoft 365 connection, and off until you switch it on, Secuno writes each Windows device's verdict to an Entra device attribute every 15 minutes. A Conditional Access policy can filter on it. Devices synced from on-premises Active Directory cannot be written, and switching it off clears what Secuno wrote.

From finding to fix

The finding and the fix in the same place.

Secuno already tells you what is wrong. Device management closes the loop: the setting that fixes it is one click away, aimed at exactly the devices that fail.

Fix with a profile

A failing CIS check, a failing compliance control or a vulnerability with a vendor workaround opens a new profile filled with the settings that make it pass, aimed at the failing devices.

Evidence that names its source

A compliance control names the profile that enforces it, with the devices that comply and since when. Only a value that passes counts: a 60-minute screen lock is no evidence for a 15-minute rule. An evidence pack for NIS2, ISO 27001 or CIS v8 lists it per control, with the exceptions and the gaps, as a PDF for the auditor.

Workarounds that stay applied

Eight of the built-in vendor workarounds can be enforced by a profile with drift correction. A risk then names the devices where a profile already covers it.

Monitors

Monitors that fix what they find.

The agent samples every minute over the monitor's duration, so a brief spike is not an alert. A breach opens one alert and, if you want, one ticket. When it recovers, Secuno notes it and closes a ticket nobody touched.

What can be monitored12
Low diskService or systemd unitCPUMemoryWindows event IDScript checkCertificate expiryTCP portPingLinux journalDevice offlineBackup too old
Device offline and backup age are judged by the platform. The rest run on the agent.

Self-healing

A monitor can run an approved script when it breaches, at most three times a day per device by default. The ticket waits 30 minutes: if the script fixed it, the alert is marked self-healed and never becomes a ticket. If the script failed, the ticket opens at once with its output.

Quiet during planned work

A breach that starts during a mute, the client's maintenance window or the first 15 minutes after a restart opens nothing unless it outlasts them. Mutes cover a client or chosen devices, all monitors or one, with a reason, for up to 30 days.

Wired into automation

Breached and recovered are automation triggers, and monitor figures (mean time to recover, self-healed alerts, checks in breach) can go on your own dashboards, with a ready-made Device monitoring template.

Software

The right software, kept that way.

Mark an app required and it stays installed. Installs run through the same queue as patches, inside the maintenance window, with three attempts per device.

Required, one-time or available

Required apps are put back if someone removes them. One-time apps are installed once. Available apps are offered on the device page for one-click install. Push app installs on chosen devices now, also across clients.

Windows and Linux, from trusted sources

winget on Windows, searchable from the console with how many of your devices have each result. On Linux apt, dnf, zypper, flatpak or snap, only from the host's own repositories and after a dry run that refuses any removal. Install an exact version and hold devices at it.

Your own packages, checked twice

Upload MSI, EXE, DEB or RPM packages up to 500 MB. Someone other than the uploader approves them. The agent downloads through a short-lived link valid only for that install, checks the hash before it runs and checks the result afterwards.

Keys and passwords

The recovery key is there when the laptop is not.

BitLocker recovery keys and local admin passwords are stored encrypted, bound to their client and device. The order of operations is designed so a failure can never leave a device locked with no key on record.

  • BitLocker is turned on only after its recovery password was escrowed. Secuno never decrypts a drive or interrupts one that is decrypting
  • The managed local admin account is created and its password rotated by the agent, which escrows the new password before it sets it. A failed upload changes nothing
  • Revealing a key or password asks for a reason, is audited and notifies the client's super users and the partner's admins
  • A revealed secret is rotated eight hours later. Rotate now does it at once
  • The Secrets view lists what is escrowed across all your clients
Control and accountability

Powerful actions, with a second pair of eyes.

Being able to manage a device means being able to break it. The risky actions need a second person, and none of them happens without an audit record.

Lock and retire

Lock locks every session on the device, with an optional message first. Retire puts back every setting Secuno changed and then decommissions the device. Escrowed keys stay until the client is offboarded.

A wipe takes two people

Wipe works on Windows workstations only. You type the hostname, and a second person with access to that client approves within 24 hours. Automation rules and the command API cannot send a lock or a wipe.

Change approvals

Switch on change approval for a client and every new, changed or deleted profile, compliance policy or app assignment, and every app push, waits for someone other than its author. The approver sees what each field and setting would change. Requests expire after seven days.

Scope tags

A technician with scope tags sees and changes only the rules and devices that carry one of them. Any other device answers not found. Devices get a tag by hand, many at once or by a rule.

Zero-touch onboarding

An onboarding profile adds tags and groups to a new device at enrolment and defines when it is ready: settings applied, required apps installed, keys escrowed, compliant. The device page shows the progress, and an enrolment token can name the profile to use.

No audit record, no action

Wipes, locks, key and password reveals, rotations and approvals are written to the audit log in the same transaction as the action. When the log cannot be written, the action is refused.

Questions

What MSPs ask first.

Does this replace Intune or Group Policy?

It does not have to. Where a GPO or MDM policy already owns a setting, Secuno reports it as managed elsewhere and does not write it. Secuno also reads your Intune policies and shows them beside its own settings, and it never changes them. Use Secuno profiles for clients without Intune and for the hardening settings your MDM does not cover, and see compliance for all of them in one place.

Can a profile lock people out?

That is what report mode, the pilot stage and the halt gate are for. Report mode checks without changing anything. A staged rollout reaches your pilot devices first and stops automatically when more than 5% of devices fail a changed setting or go silent.

What happens when I remove a setting from a profile?

By default the agent puts back the value from before Secuno changed it. You can choose per profile to leave devices as they are instead, which is what profiles created before 31 October 2026 do. BitLocker, managed local admin passwords, check-only settings and a setting a compliance policy still checks are never reverted: decrypting a disk is not something to do automatically.

Can someone wipe a device by mistake?

It takes two people. The person asking types the hostname, and a second person with access to that client has to approve within 24 hours. Wipe exists for Windows workstations only, and automation rules cannot send it.

Which platforms are covered?

Windows and Linux. macOS is not supported. Not part of this version: wiping a Linux device, enabling a Linux host firewall, escrowing LUKS keys and MSIX packages. The Linux host firewall is checked, not written.

Where are the BitLocker keys kept?

Encrypted on the Secuno platform in the EU, each key bound to its client, device and volume. Every reveal is audited, notifies the client's and partner's admins, and leads to a rotation eight hours later.

What does it need on the device?

The Secuno agent you already run: Windows agent 0.63 or Linux agent 0.55 and later. Reverting settings and remote lock, retire and wipe need Windows 0.64 or Linux 0.56. No second agent and no separate MDM enrolment. A device on an older agent keeps its protection but is not managed until its agent is updated.

Next step

Manage the devices you already protect.

One agent, one console, one place where a finding turns into the setting that fixes it. We will show it on a fleet like yours.