CVE-2026-68903: LLMKube · Arbitrary Host File Read via Model.spec.source hostPath Injection
Affects github.com/defilantech/llmkubeSource GCVE Database
Not known to be exploited · CVSS 7.7 High : upgrade to 0.9.0 soon.
Data as of 19 Nov 2026, 00:21 UTC
Thin evidence
One scorer or none| CNA (GitHub_M) | CVSS 3.1 | 7.7 |
|---|
- Patch
- Available – fixed in 0.9.0
CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:C/C:H/I:N/A:NExploitation evidence
Disclosed: evidence found.
CVE published 18 Nov 2026
Detectable: no evidence on record.
No public nuclei template indexed
Weaponized: no evidence on record.
No public exploit code indexed in Metasploit or Exploit-DB
Exploited in the wildExploited: no evidence on record.
Not in CISA KEV and no source reports exploitation
Observed by OffSeq MirageSeen by Mirage: no evidence on record.
No attack traffic recorded by OffSeq Mirage sensors
A missing step is not evidence of safety. Sources: cve.org (opens in a new tab), the Metasploit, Exploit-DB and nuclei-templates index as of 18 Nov 2026.
What to do
Upgrade github.com/defilantech/llmkube to 0.9.0.
Affected software
| Product | Affected | Fixed in |
|---|---|---|
| Gogithub.com/defilantech/llmkube | Affected
| Fixed in
|
Check your own hosts for these packages with threat-finder (opens in a new tab), our free open-source scanner.
AI-suggested
A patch is available in LLMKube version 0.9.0 that fixes this vulnerability. Users should upgrade to version 0.9.0 or later. Since model caching, which enables the vulnerable hostPath volume mount, is enabled by default, upgrading is critical to prevent exploitation. No vendor advisory states that no action is required or that the issue is already mitigated without patching.
Description
LLMKube versions prior to 0.9.0 contain a vulnerability in the ModelReconciler component where the Model.spec.source field accepts arbitrary absolute paths and file:// URIs without path allowlisting. This allows users with create/update permissions on Model and InferenceService custom resources to read arbitrary files accessible to the controller pod or on the node hosting the inference pod, including sensitive files like Kubernetes service-account tokens and kubelet configuration. Model caching, which triggers the vulnerable hostPath volume mount, is enabled by default. The vulnerability has a CVSS 3.1 base score of 7.7 (High).
References
- github.com/defilantech/LLMKube/security/advisories/GHSA-jw3m-8q7m-f35r (opens in a new tab)
- github.com/defilantech/LLMKube/pull/981 (opens in a new tab)
- github.com/defilantech/LLMKube/commit/ec2647fd4557962a971a1bc9435790b8e77ce961 (opens in a new tab)
- github.com/defilantech/LLMKube (opens in a new tab)
- github.com/defilantech/LLMKube/releases/tag/v0.9.0 (opens in a new tab)