Device Trust
Last modified on July 31, 2024
This feature is part of the Enterprise plan. If it is not enabled for your organization, please contact StrongDM at the StrongDM Help Center.
Overview
Device Trust is a security mode that enables your organization to configure StrongDM to work with endpoint management software, such as CrowdStrike and SentinelOne.
When running in Device Trust mode, the StrongDM client provides signals on the host machine’s Device Trust status when evaluating policies that require it. The information that is provided to the policy evaluation indicates whether the machine hosting the client has an endpoint agent running and if it has flagged any vulnerabilities. If any of these conditions are violated, the client can be logged out and any active resource connections are severed.
StrongDM admins can enable Device Trust for all users, including service accounts. Moreover, admins can specify users and roles to be excluded from Device Trust enforcement. This page describes how to set up Device Trust in the Admin UI.
Prerequisites
- Enablement requires that your organization have software from a supported Device Trust provider running on user workstations.
- You must have administrative access to your organization’s Device Trust provider account.
- You must enact policies that use Device Trust as a condition of access in order to trigger Device Trust checks.
Admin UI Configuration
- In the Admin UI, go to Settings > Device Trust.
- Click the lock to change the settings.
- For Enable Device Trust for Your Organization?, select Enabled to enable it. Then click Save.
- For Default User Enforcement you can choose whether you wish for users to have Device Trust enforced on them by default, or to be exempt by default.
- Consider which user(s) and/or role(s), if any, that you want to be excluded from Device Trust enforcement. Before fully activating Device Trust, you can make exceptions for specific users and roles so that Device Trust is enforced for all users and roles except the ones you exclude. To make an exception for a user, go to the user’s Settings tab and check the relevant option for Device Trust Enforcement. To make an exception for a role, go to the role’s Settings tab.
- Go back to Settings > Device Trust to configure the remaining settings.
- For Provider, choose your endpoint management software provider (for example, CrowdStrike or SentinelOne).
- Complete the remaining settings for your selected provider. In these settings, “agent” refers to the endpoint agent that is installed on the user’s workstation. The agent monitors the user workstation’s posture and assesses whether the given workstation is in a positive or negative integrity state.
- Allow service accounts to run without the agent allows you to exempt service account machines that are not enrolled with your Device Trust provider.
- Save when you’re done.
- Lastly, ensure that policies are configured to forbid or permit resource access based on the Device Trust status of a user’s workstation.
Device Trust provider settings and requirements
Cisco Duo
Settings
Setting | Requirement | Description |
---|---|---|
Integration Key | Required | Integration key copied from your Duo application settings in Duo Admin |
Management URL | Required | Duo Admin API hostname/endpoint (for example, api-a12b3-c45.duosecurity.com ) copied from your Duo application settings in Duo Admin |
Secret Key | Required | Secret key copied from your Duo application settings in Duo Admin |
Additional requirements for Duo
In order for StrongDM to communicate with Duo, your Duo Admin application must have the “Grant read resource” permission enabled. You can check that your application has the correct permissions in Duo Admin > Dashboard > Applications.
CrowdStrike
Settings
Setting | Requirement | Description |
---|---|---|
Base URL | Required | CrowdStrike base address (for example, https://your-cloud-region.crowdstrike.com ) |
Client ID | Required | CrowdStrike client ID |
Client Secret | Required | CrowdStrike client secret |
Member CID | Optional | CrowdStrike customer identification (CID), which is found on the sensor download page of the CrowdStrike Console |
Provider | Required | Select CrowdStrike |
Score | Required | Numeric value, from 1 to 100, that indicates the security posture for the host |
Additional requirements for CrowdStrike
The minimum scopes required when creating your CrowdStrike credentials are:
- Hosts (Read)
- Zero Trust Assessment (Read)
Microsoft Defender
Settings
Setting | Requirement | Description |
---|---|---|
Client ID | Required | Client ID for your Microsoft Defender app, available on the app’s Overview page |
Client Secret | Required | Client Secret for your Microsoft Defender app, copied after creating it in your app by selecting Add a certificate or secret in the Client credentials section |
Max Risk Score | Required | “None”, “Informational”, “Low”, “Medium”, “High” |
Tenant ID | Required | Tenant ID for your Microsoft Defender app, available on the app’s Overview page |
Additional requirements for Defender
The application requires the Machine.Read.All
permission added within the WindowsDefenderATP API. To add the permission, follow these steps.
- Go to API permissions > Add a permission.
- In the APIs my organization uses table, search for and select WindowsDefenderATP.
- On the resulting screen, search for and select the
Machine.Read.All
permission. This action may require approval from an administrator.
SentinelOne
Settings
Setting | Requirement | Description |
---|---|---|
API Token | Required | SentinelOne API token, which can be generated in the SentinelOne management console in the user settings |
Management URL | Required | SentinelOne Management URL (for example, https://example-management-url.sentinelone.net/ ) |
Provider | Required | Select SentinelOne |
Additional requirements for SentinelOne
SentinelOne API calls require authentication, and SentinelOne’s recommended authentication is API token (that is, ApiToken
). Your SentinelOne credentials need to have view permissions for the app(s) you want to monitor. API tokens are generated in the SentinelOne Management Console or your API request, and each token is valid for six months. Because of this expiration, you must rotate/regenerate your API token every six months, if SentinelOne is your Device Trust provider type. You can see your token’s expiration date when viewing your user account in the SentinelOne Management Console.
For information, please refer to SentinelOne API documentation.
User Administration
Particular users can have a Device Trust setting that is explicit and independent of the organization-wide controls. The values available when editing a user’s profile are:
- Default: Device Trust is enforced for this user if the organization’s global settings require Device Trust to be enforced.
- Exempt: Device Trust is not enforced for this user, regardless of global settings.
- Required: Device Trust is enforced for this user, regardless of global settings.
User Experience
When Device Trust is enabled, administrators can use policies to check the device status of any users with it enabled. This check can then be used to validate whether their device posture as reported by the agent is still good. If their device posture is not acceptable, the policies can force further actions such as MFA re-authentication, request to log the user’s reasons for particular actions, or forcibly log out the user.
If the user is using the CLI and is forcibly logged out, all connections are severed.