Authenticated command execution via unsanitized model URLs
Published Apr 6, 2026 · Updated Apr 6, 2026
OS command injection in KubeAI before 0.23.2 allows remote authenticated users to execute commands inside model server pods. The ollamaStartupProbeScript function interpolates unsanitized ref and modelParam URL components into a shell string passed to bash -c. Create or update permission for Model custom resources is required, granting command execution in the model pod and access to its environment, mounted secrets, and service account token.
Summary
What happened
OS command injection in KubeAI before 0.23.2 allows remote authenticated users to execute commands inside model server pods. The ollamaStartupProbeScript function interpolates unsanitized ref and modelParam URL components into a shell string passed to bash -c. Create or update permission for Model custom resources is required, granting command execution in the model pod and access to its environment, mounted secrets, and service account token.
The record
- CVE
- CVE-2026-34940
- Published
- Apr 6, 2026
- Updated
- Apr 6, 2026
- Vendor
- KubeAI Project
- Product
- KubeAI
- Classifications
- CWE-78, T1059.004
- Attack vector
- network
- Privileges
- authenticated
Timeline
How it unfolded
- Apr 6, 2026CVE publishedPublication date reported by the CVE source.
- Apr 6, 2026Record updatedLatest update available in the CVE record.
Exploitability
Present is not the same as exploitable
Compare your product and version with the public record. A matching version still requires validation against your environment.
Is a vulnerable build present?
Compare these published version ranges with your installed build and any vendor patches.
- Affected versionversion=< 0.23.2
What conditions does exploitation require?
What is affected?
Attacks
What attackers are doing with it
Daily unique IPs observed by Shadowserver honeypots for known exploited vulnerabilities (KEVs). Missing observations do not establish an absence of attacks.
Weakness, pattern, technique
Public exploit references
- Model custom-resource command-injection proof of conceptproof of concept · demonstrated
Labels summarize the accepted research assessment. They do not indicate a test against your environment.
Technologies
Your stack
See the directory against your own environment.
Your stack
Check the software in your environment
Book a demo to see how Hinoki identifies affected software and validates exploitability in your environment.
Book a demo