Asset-discovery services often receive permission to scan broad address ranges and collect detailed endpoint data. A denial-of-service flaw in that service can hide the very inventory used to investigate an incident. Tanium security bulletin TAN-2026-124 describes CVE-2026-131653, an improper input-validation issue in Discover.

The bulletin rates the vulnerability medium with a CVSS v3 base score of 4.8. An authenticated Tanium user with the Discover Profile Write permission can submit malformed profile input and cause a denial of service against Discover. The vector requires network access and low privileges but no user interaction; the stated impact is limited to availability.

The permission is the attack boundary

This is not an unauthenticated remote crash. The attacker needs a Tanium identity and a write permission on discovery profiles. That makes the vulnerability relevant to delegated administration, service accounts, and role design.

An account that can edit profiles is already trusted to change what the organization scans and how results are collected. The advisory shows that this authority also carries an availability risk for the Discover service. Teams should not distribute profile-write permissions to every operator who can view inventory or run a saved query.

Affected versions are listed by release family. The 2025H1 line is affected from v4.10 through v4.10.209, the 2025H2 line from v4.15 through v4.15.224, and the 2026H1 line from v4.19 through v4.19.145. The fixed baselines are update 28 at v4.10.209, update 18 at v4.15.224, and update 11 at v4.19.145.

The bulletin lists no workaround. Patching is therefore the primary remediation, but permission review and detection remain useful while the update is being scheduled.

Inventory the vulnerable management path

Check the Discover server version and each management component that accepts profile changes. Confirm whether the affected version is still supported in the deployed release family and whether the fixed baseline is actually installed rather than merely available.

Map every principal with Discover Profile Write. Separate human administrators from automation accounts, compare the role assignment with the owner of each profile, and remove write rights from accounts that only need to read results. Rotate credentials for any unexpectedly broad service account after the review.

Inspect audit records for profile creation and update requests before a service interruption. Look for malformed fields, repeated submissions from the same account, requests that do not match an approved change window, and a sudden absence of Discover telemetry after a profile update. Correlate the event with authentication logs and endpoint management activity.

Keep discovery available during an incident

The availability impact is more serious when Discover is the only map of the environment. Maintain a documented export of critical asset ownership, network ranges, and endpoint coverage that does not depend on the same service being online. Test that another team can read the inventory during a degraded-management exercise.

Limit network access to the Discover API and administrative interfaces. An authenticated attacker still needs a route to the service. Apply the rule that a low-privilege account should not have an unconstrained management path to the monitoring plane.

After the update, retest profile creation with invalid and boundary-case values in a non-production environment. The vendor release should handle malformed input without stopping the service, but only a controlled test confirms that the deployed package behaves as expected.

Verification after the patch

Record the old and new build identifiers, the update timestamp, and the owner who approved the change. Confirm that the service restarts cleanly and that existing profiles and scan schedules remain intact. Review logs for a recurrence of the malformed request pattern.

The vulnerability is medium severity, but it deserves a place in the identity and continuity review. A profile editor is not just a configuration screen. It is a control point for discovery coverage, and losing that control point while an incident is in progress can cost more than the service outage itself.

References