Skip to main content

Identity Alerts

If you already run an identity threat product, OrbisID can read its alerts and count them in the Threat Activity part of each account's risk score, so a privileged account that your security tools are worried about scores higher, without an Endpoint Sensor on every host. Three sources are supported:

  • Microsoft Defender for Identity, through the Microsoft Graph security alerts API.
  • Microsoft Entra ID Protection, through the Microsoft Graph risk detections API.
  • CrowdStrike Falcon Identity Protection, through the Falcon Alerts API.

This is available in the Pro and Enterprise editions. The Administrator role turns each source on under Administration > Integrations > XDR Identity Alerts. Both are off until you configure them. This is the recommended way to feed threat activity into OrbisID: Endpoint Sensors are optional, a fallback for estates without an identity threat product, and keep working alongside these sources if you have them.

How alerts are counted​

  • OrbisID fetches alerts every ten minutes (and when you choose Fetch now), asking each product only for alerts changed since the last poll. When alerts arrived, risk scores are recalculated.
  • Each alert is matched to OrbisID accounts by the name of the affected user that the product reports. For Defender for Identity that is the user principal name and the sign-in name (and DOMAIN\name) from the alert's user evidence; an account matches when its name equals one of them, ignoring case. If none matches, the part of an address before the @ is tried, but only when exactly one account has that name. An account has to have been scanned to be matched. An alert about a user with no OrbisID account is kept and listed on the Identity Alerts page, but counts in no score.
  • An open alert (Defender status new or in progress; CrowdStrike status not closed) raises the account's Threat Activity score to the alert's weight: the highest weight among all its open alerts and any open Threat Detections decides the score. The factor list names each source, for example "2 open Microsoft Defender for Identity alerts, highest weight 85%".
  • A resolved alert, or one Defender classifies as a false positive, stops counting. An open alert nobody closes stops counting 30 days after its last update (you can change this per source, 1 to 365 days).
  • Weights: for Defender for Identity, informational 15, low 35, medium 60, high 85, unknown 30. For CrowdStrike, the alert's own severity (0 to 100).

An account's details show an Identity alerts section listing its alerts, their source, severity, whether they still count, and a link to the alert in the product.

Microsoft Defender for Identity​

  1. In Microsoft Entra ID, register an application (or reuse one) and create a client secret.
  2. Under API permissions, add the Microsoft Graph application permission SecurityAlert.Read.All and grant admin consent. This is the least privileged permission for listing alerts.
  3. In OrbisID, open Administration > Integrations > XDR Identity Alerts → Microsoft Defender for Identity and enter the directory (tenant) ID, the application (client) ID and the secret, turn it on, and choose Test sign-in, then Fetch now.

OrbisID reads only alerts whose service source is Defender for Identity (serviceSource eq 'microsoftDefenderForIdentity'). The public Microsoft cloud is supported.

Microsoft Entra ID Protection​

Entra ID Protection raises risk detections about users: leaked credentials, impossible travel, password spray, anonymous or malicious IP addresses and similar. It needs Microsoft Entra ID P1 or P2 in the tenant.

  1. In Microsoft Entra ID, register an application (or reuse the one for Defender for Identity) and create a client secret.
  2. Under API permissions, add the Microsoft Graph application permission IdentityRiskEvent.Read.All and grant admin consent.
  3. In OrbisID, open Administration > Integrations > XDR Identity Alerts → Microsoft Entra ID Protection and enter the directory (tenant) ID, the application (client) ID and the secret, turn it on, and choose Test sign-in, then Fetch now.

Each detection is matched by the user's principal name. Its weight follows the detection's risk level (none 15, low 35, medium 60, high 85, hidden 30). A detection counts while its risk state is at risk, confirmed compromised or none; once an administrator or Microsoft marks it remediated, dismissed or confirmed safe it stops counting. The detection's link back to the portal is kept when Microsoft supplies one.

CrowdStrike Falcon Identity Protection​

  1. In the Falcon console, create an API client with the Alerts: Read scope and note its client ID and secret.
  2. In OrbisID, open Administration > Integrations > XDR Identity Alerts → CrowdStrike Falcon Identity Protection, enter the API base URL of your Falcon cloud (for example https://api.crowdstrike.com; it must end in .crowdstrike.com), the client ID and the secret, turn it on, and choose Test sign-in, then Fetch now.

OrbisID asks for alerts with product:'idp' that changed since the last poll, then retrieves them by composite ID. The Falcon alert includes the user it concerns under a field name that can vary; OrbisID looks for user_principal_name, user_name, username, account_name, email_address and email in that order. If Fetch now reports alerts as unmatched when you expect matches, the page lists them with the user OrbisID found (or none), which shows whether the field name needs adding.

Secrets and data​

The secrets are stored encrypted and are never shown again; leave the secret blank when saving to keep the saved one. OrbisID stores each alert's title, severity, status, category, link, the user it named, and its top-level scalar fields, so a response can be checked. Nothing is written back to either product.

Microsoft and CrowdStrike documentation​

Entra ID Protection follows Microsoft Learn: List riskDetections (endpoint, IdentityRiskEvent.Read.All, Entra ID P1 or P2, $top up to 500) and the riskDetection resource type (properties, risk levels and states). Microsoft's page names $filter as supported without listing the filterable properties; OrbisID filters on lastUpdatedDateTime and, if Graph refuses, reads without the filter and keeps only what changed.

Defender for Identity follows Microsoft Learn: List alerts_v2 (endpoint, SecurityAlert.Read.All, $filter on serviceSource and lastUpdateDateTime, @odata.nextLink), alert resource type, userEvidence, userAccount and the security enum values.

CrowdStrike follows the Alerts collection of the Falcon developer reference (GET /alerts/queries/alerts/v2 with filter, limit, offset and sort) and FalconPy's endpoint definitions (POST /alerts/entities/alerts/v2, filter fields product and updated_timestamp).