Skip to main content

Key Risk Indicators (KRIs)

KRIs are metrics that quantify your privileged access risk posture. Each KRI measures a specific aspect of risk and uses RAG (Red/Amber/Green) thresholds to indicate severity.

KRI Overview​

Navigate to KRIs to see all current indicator values.

KRI Trends

Each KRI card shows:

ElementDescription
NameWhat is being measured
Current ValueLatest count or percentage
RAG StatusGreen (within tolerance), Amber (warning), or Red (critical)
ChangeIncrease or decrease since the last snapshot

Built-in KRIs​

KRI cards are shown in the order below. The Community column shows which 4 KRIs are included in the free edition — all others require Pro or Enterprise (see Licensing).

KRIDescriptionCommunity
Privileged Without OwnerEnabled privileged accounts with no linked identityYes
Standing PrivilegesAll enabled accounts with privileged entitlements (always-on, not just-in-time)Yes
Not in PAM ToolPrivileged accounts not linked to a PAM inventory entryYes
Shared/Non-Human Privileged AccountsNon-human (shared/service) accounts with privileged entitlementsPro+
Inactive Identity with Privileged AccountPrivileged accounts linked to inactive or out-of-date identitiesPro+
High-Risk Unmanaged Privileged AccountsPrivileged accounts on High/Critical-risk systems not in PAMPro+
Stale Passwords on Privileged AccountsPrivileged accounts with passwords older than the threshold agePro+
Dormant Privileged AccountsPrivileged accounts with no recorded use (the latest last-used date from any source) within the threshold periodPro+
Disabled Accounts in PAMDisabled accounts that still exist in the PAM inventoryPro+
Account-Identity Type MismatchAccounts where the account type doesn't match the linked identity's typePro+
Unlinked AccountsAll accounts (privileged or not) not linked to any identityYes
Systems Never ScannedActive systems that have never been scannedPro+
Systems with Stale ScansActive systems not scanned within the threshold periodPro+
Systems Without CredentialsActive systems with no configured credentialsPro+
Privileged Services with Non-Privileged AccountPrivileged services running as a non-privileged accountPro+
Privileged Services Without AccountPrivileged services with no linked run-as accountPro+
Scan Failure RatePercentage of scans that failed or partially completed within the periodPro+
Unclassified EntitlementsEntitlements not yet classified as privileged or non-privilegedPro+
Privileged Account ConcentrationSystems exceeding the privileged account concentration thresholdPro+
PAM Coverage RatioPercentage of privileged accounts managed in your PAM toolPro+
Open Threat DetectionsUnresolved Threat Detections currently in Open statusPro+
High-Confidence Threat DetectionsOpen Threat Detections with an ML-adjusted confidence score above 0.80Pro+
Accounts with Active Threat DetectionsDistinct privileged accounts with at least one open Threat DetectionPro+
Confirmed Threat Detection RatePercentage of resolved Threat Detections confirmed as true positivesPro+
Unacknowledged High-Confidence ThreatsOpen high-confidence Threat Detections left unacknowledged past the threshold agePro+
Offline Monitor AgentsEndpoint Sensors that stopped heartbeating and were marked OfflinePro+
Systems with Overdue Attestation RequestsSystems with an attestation request that expired without being submitted or cancelledPro+
Long-Lived CredentialsAccounts with an active key, secret, certificate or token last rotated (or created) more than the threshold number of days ago (default 365)Pro+
Expired Credentials Still ActiveAccounts with a credential past its expiry date that the source still reports as activePro+
Credentials Never RotatedAccounts with an active credential created more than the threshold number of days ago (default 180) that shows no rotation since it was createdPro+
Non-Human Accounts Without OwnerEnabled non-human accounts (service accounts, API keys, application identities) not linked to an owner identityPro+
Orphaned Non-Human AccountsEnabled non-human accounts whose owner identity is inactive, has not started or has passed its end datePro+
Credential Policy BreachesAccounts with an active credential over the maximum age or rotation interval of a Credential Policy rule. Zero until a rule existsPro+
Secrets Outside the VaultCI/CD organisations and repositories holding a secret that no vaulted account explains by name (see Secrets Outside the Vault)Pro+
Systems Overdue for Scheduled AttestationSystems on a scheduled attestation cadence that haven't completed one within their configured periodPro+
Privileged Access Attestation Completion RatePercentage of attestation requests submitted rather than left to expirePro+

The two ownership KRIs apply to every enabled account classified as non-human, privileged or not. The owner of a non-human account is the identity it is linked to, the same link a human account uses; link it from the account's details or with bulk linking on the Accounts page. Accounts not yet classified as human or non-human are not counted.

The threat detection KRIs (Open Threat Detections, High-Confidence Threat Detections, Accounts with Active Threat Detections, Confirmed Threat Detection Rate and Unacknowledged High-Confidence Threats) and Offline Monitor Agents depend on Endpoint Sensors or an identity alert source. Where none is set up they show Not set up instead of a green zero, because nothing is watching rather than nothing happening: the threat detection KRIs once any sensor, identity alert source or earlier detection exists, and Offline Monitor Agents once a sensor is enrolled. Control monitoring treats them the same way and shows no data for the controls that rely on them.

The three credential KRIs count accounts, not credentials, using the credential metadata from scans (AWS access keys, Entra ID secrets and certificates, Google Cloud service account keys, GitHub personal access tokens). A credential whose date the system does not report is never counted, because unknown is not a breach. Where a system only records one date for creation and rotation (AWS access keys), Credentials Never Rotated and Long-Lived Credentials flag the same credentials once both thresholds are passed. A credential that is replaced by a new one is simply the new credential, so replacing a key clears the account.

KRI Definitions​

Navigate to KRIs and view the definitions to see or edit the configuration for each KRI.

Each definition has:

FieldDescription
CodeUnique identifier (e.g., PRIVILEGED_WITHOUT_OWNER)
NameDisplay name
DescriptionWhat the KRI measures and why it matters
Green ThresholdValues at or below this are healthy
Amber ThresholdValues between Green and Red are a warning
Red ThresholdValues at or above this are critical
EnabledWhether the KRI is active
Create findingsWhether each item the KRI flags becomes a finding (Pro and Enterprise; only KRIs that flag individual items)
Finding severityThe default severity of those findings
note

Thresholds are inclusive. For example, if Green = 5, Amber = 15, and Red = 16, then a value of 5 is Green, 15 is Amber, and 16 is Red.

Editing Thresholds​

  1. Click on a KRI definition
  2. Adjust the Green, Amber, and Red threshold values
  3. Click Save

Threshold changes take effect immediately on the dashboard and KRI pages.

KRI Snapshots​

Requires Pro or Enterprise edition.

Snapshots capture a point-in-time record of all KRI values. They are used to build trend charts and track progress over time.

Automatic Snapshots​

A snapshot is taken automatically each time a scan completes.

Manual Snapshots​

  1. Navigate to KRIs > Snapshots
  2. Click Take Snapshot

Viewing Snapshots​

The snapshots page shows a table of all historical snapshots with their timestamp and KRI values. Click a snapshot to see its full breakdown.

Requires Pro or Enterprise edition.

Navigate to KRIs > Trends to view KRI values over time as a chart.

  • Select one or more KRIs to display
  • Choose a time range
  • The chart background uses RAG colouring to show threshold bands

This makes it easy to spot whether your risk posture is improving or deteriorating.

KRI Exceptions​

Requires Enterprise edition.

Exceptions allow you to formally accept the risk of specific items (accounts, systems, entitlements or services) for a KRI. This is useful for items that have been reviewed and accepted as a known risk, such as a break-glass account that must stay outside the PAM tool.

An exception needs three things: a justification, an expiry date, and an approver who is not the person asking. It has no effect until the approver approves it.

How an exception works​

StatusWhat it means
Pending approvalRequested; waiting for the chosen approver. The item still counts towards the KRI.
ActiveThe approver approved it and its expiry date has not passed. The item is left out of the KRI's adjusted value, and its finding is Accepted Risk.
RejectedThe approver said no, and why. The item still counts, and its finding stays open.
ExpiredAn approved exception past its expiry date. The item counts towards the KRI again and its finding reopens at the next refresh, at the latest the nightly one.

Exceptions that were created before approval was introduced stay active as they were, with no approver, and possibly with no expiry date. To change an exception's reason or expiry, an Administrator deletes it and a new one is requested.

Requesting an Exception​

  1. Navigate to KRIs > Exceptions
  2. Click Request Exception
  3. Fill in:
FieldRequiredDescription
KRIYesWhich KRI this exception applies to
Entity Type and EntityYesThe accounts, systems, entitlements or services being excepted. Choose several to send one request for each
ReferenceNoA change or ticket number
JustificationYesWhy the risk is acceptable
Expiry DateYesThe last day the exception applies; it must be in the future
ApproverYesAny other active IAM Governance Manager or Administrator. Someone with no email address on their account is marked, because they would not be notified. An Administrator can also choose themselves; see below
  1. Click Request approval

You can also request one from a finding with Request risk acceptance. While an exception for the same KRI and item is pending or active, you cannot request another. Once it has been rejected, has expired or has been deleted, you can; the new request replaces the old record.

Approving or Rejecting​

The approver is emailed with a link to the Exceptions page. On that page, Awaiting my decision shows the requests waiting for you, and each row has Approve and Reject. The same buttons appear on the finding itself. Approving can carry a comment; rejecting needs one, so the requester knows why.

Only the approver chosen for a request can decide it, and never the person who made it. The requester is emailed the decision.

Administrators approving their own request. An Administrator can choose themselves as the approver, for example on an installation with only one Administrator. A warning appears and they must tick a box to confirm they are approving their own request. The exception then takes effect straight away, with no email, and the audit log records it as a self-approval. Managers cannot do this.

Managing Exceptions​

  • The person who made a pending request can edit its justification, reference and expiry date, or withdraw it by deleting it.
  • Once decided, only the reference can be edited. A longer expiry or a different reason needs a new request (after an Administrator deletes the current exception, or once it has expired), so nobody can extend an approval by editing it.
  • An Administrator can delete any exception. This revokes it: the item counts towards the KRI again and its finding reopens at the next refresh.
  • The requester is emailed a reminder shortly before an approved exception expires (by default 14 days before, once). Administrators set this in Administration > Settings > Findings & Risk > KRI Exceptions; 0 turns it off.
  • The wording of the four emails (request, approved, rejected, expiring) can be changed in Administration > Integrations > Email Templates.
  • Every request, decision, edit and deletion is recorded in the audit log.