CVE-2026-34940 is an OS command injection vulnerability (CWE-78) in kubeai-project's KubeAI product affecting versions before 0.23.2. The vulnerability arises because the ollamaStartupProbeScript() function uses fmt.Sprintf to build a shell command string incorporating unsanitized input from model URL components (ref, modelParam). This command is executed via 'bash -c' as a Kubernetes startup probe, allowing an attacker who can create or update Model custom resources to execute arbitrary shell commands within the model server pods. The issue is resolved in version 0.23.2.

Successful exploitation allows an attacker with Model custom resource creation or update privileges to execute arbitrary shell commands inside model server pods. This can lead to complete compromise of the affected pod's environment, including potential data exposure and disruption of service. The CVSS score is 8.7 (high), reflecting network attack vector, low attack complexity, required privileges, no user interaction, and high confidentiality and integrity impact without availability impact.

Upgrade KubeAI to version 0.23.2 or later, where this vulnerability is fixed. Since no official patch or workaround is detailed beyond this version update, applying the upgrade is the recommended remediation. Patch status is not explicitly confirmed beyond the fix in 0.23.2, so users should verify with the vendor advisory for the latest guidance.