CVE-2026-110697

Quay is a system that stores and manages container images for software teams, and each customer or team gets a separate 'namespace' with its own secret access tokens (robot accounts) used like passwords for automated systems. This flaw is like a filing cabinet where entering a slightly odd search term accidentally pulls up other customers' private folders — someone could craft a namespace name that tricks the search into revealing another tenant's secret tokens, which they could then use to access or tamper with that tenant's container images.

This matters to you if you run Red Hat Quay 3, Red Hat Quay 5, or Red Hat OpenShift Update Service, particularly in a shared or multi-tenant environment where different teams or customers have their own namespaces. It's a real risk whenever an attacker can create or control a namespace on your Quay instance — the attack requires no special privileges beyond that and no interaction from a victim.

There is no confirmed exploitation in the wild and no public exploit code circulating, so this is not an active emergency today, but the flaw is severe because it lets one tenant silently steal another tenant's credentials with minimal effort. Treat it as a priority to address soon — restrict namespace creation and watch for a vendor fix — rather than something to ignore.

Confirm exposure by identifying whether your deployment runs Red Hat Quay 3, Red Hat Quay 5, or Red Hat OpenShift Update Service, and whether namespace/organization creation is open to any authenticated user (multi-tenant risk) versus restricted to trusted admins. No fixed version has been published by Red Hat at this time, so there is no upgrade path to recommend yet — monitor Red Hat's security advisories (errata) for this CVE and apply the patch as soon as it ships. As a temporary workaround, restrict namespace/organization creation to trusted administrators only, and audit existing namespace names for unusual underscore or wildcard characters that could indicate attempted exploitation. Rotate robot account tokens for any namespace that may have been queryable by untrusted users, and review Quay access logs for anomalous robot-listing API calls. No CISA KEV listing exists, so there is no federal remediation deadline; prioritize based on your own multi-tenancy exposure.

CVSS 9.9 (Critical) reflects network-exploitable (AV:N), low complexity (AC:L), requiring only low privileges (PR:L) and no user interaction, with high impact to confidentiality, integrity, and availability across a changed scope (S:C) — consistent with cross-tenant credential disclosure. The weakness is CWE-89 (SQL Injection), specifically unescaped LIKE-pattern wildcard metacharacters (underscore) in namespace-based robot account queries, enabling enumeration/retrieval of other tenants' decrypted authentication tokens. Not listed in CISA KEV; no public exploit code is known to exist, and predicted near-term exploitation likelihood is low with a flat trend. No patch is currently published — track Red Hat advisories for remediation.