OpenShift Console CVE-2026-50237

The attacker must be a low-privileged namespace tenant (PR:L per CVSS, confirmed by CVSS vector) with RBAC permission to create or modify ProjectHelmChartRepository custom resources within their namespace - this is the specific API access gate that enables URL injection. The OpenShift Console Helm catalog proxy feature must be active. For the full privilege escalation outcome, a cluster administrator must install a Helm chart from the tenant-poisoned catalog, introducing a hard dependency on admin action (social engineering or inadvertent installation from a listing the admin believes is trusted). The SSRF component alone - reaching internal network addresses or exfiltrating internal responses - does not require admin interaction and is exploitable whenever ProjectHelmChartRepository write access is granted to the tenant. Exploitation is bounded by: (1) requiring authenticated low-privileged cluster access, excluding unauthenticated external attackers; (2) the complete privilege escalation path requiring admin-mediated chart installation; (3) the console pod's network reachability, which determines what internal addresses are accessible via the SSRF.

The NVD/Red Hat CVSS base score of 7.4 (AV:N/AC:L/PR:L/UI:N/S:C/C:L/I:L/A:L) captures the SSRF trigger in isolation - a low-privileged tenant sending an arbitrary URL that the console fetches server-side - but does not fully reflect the complexity of the complete privilege escalation path. The CVSS UI:N rating is in tension with the description's explicit requirement for admin-mediated chart installation; the full attack chain requires an administrator to install the poisoned chart, which introduces a meaningful dependency on admin action not accounted for by UI:N. Real-world exploitability of the complete privilege escalation scenario is therefore lower than the CVSS score suggests. The SSRF component alone - bypassing egress to reach internal services or exfiltrate data via server-side requests - is exploitable without admin involvement and represents the more immediately actionable risk. No EPSS data was provided in the available intelligence, no CISA KEV listing exists, and no public POC has been identified, all of which further moderate near-term exploitation risk. The Scope:Changed (S:C) rating is well-grounded: the console pod's requests originate outside the tenant's namespace, crossing the namespace security boundary and potentially reaching cluster-internal infrastructure.

A low-privileged namespace tenant with ProjectHelmChartRepository write access creates a custom resource pointing to an attacker-controlled server hosting a poisoned Helm chart index. The OpenShift Console pod fetches the URL server-side - bypassing the tenant's own egress restrictions - allowing the attacker to probe internal cluster services or have the console retrieve malicious chart metadata. If the attacker then convinces or waits for a cluster administrator to install a chart from the poisoned catalog, the malicious chart payload executes with elevated cluster privileges, achieving privilege escalation; no public POC has been identified to date.