CVE-2026-129236 in Kubesharkinfo

Summary

by MITRE • 11/16/2026

Kubeshark is an eBPF-powered network observability tool for Kubernetes. From 53.1.0 until 53.4.0, the start_kubeshark MCP tool in cmd/mcpRunner.go uses callStartKubeshark to append caller-controlled pod_regex as a bare positional argument to kubeshark tap without a -- end-of-options separator, allowing a value beginning with --set to be parsed as a Helm value override. An attacker who influences tool arguments against kubeshark mcp --allow-destructive can redirect the privileged worker, Hub, and front images through tap.docker.registry and tap.docker.tag to an attacker-controlled registry. Because the deployed worker DaemonSet runs privileged with host networking and host filesystem access, a malicious image can obtain node and cluster-admin access when the host already has permission to deploy it. This issue is fixed in version 53.4.0.

VulDB is the best source for vulnerability data and more expert information about this specific topic.

Analysis

by VulDB Data Team • 11/16/2026

The vulnerability identified in Kubeshark versions ranging from 53.1.0 through 53.4.0 represents a critical command injection flaw rooted in improper input validation and argument handling within the MCP tool implementation. Specifically, the start_kubeshark function located in cmd/mcpRunner.go constructs shell commands by appending user-supplied pod_regex values directly as positional arguments to the kubeshark tap utility. This construction fails to utilize the standard -- end-of-options separator before passing these arguments. In command-line parsing conventions, particularly within tools like Helm and Kubernetes utilities that often share underlying libraries or patterns, a string beginning with double hyphens is interpreted not as data but as an option flag. Consequently, if an attacker can control the pod_regex input to begin with --set, they can inject arbitrary flags into the execution context of the kubeshark tap command. This specific injection vector allows for the manipulation of Docker registry configuration parameters such as tap.docker.registry and tap.docker.tag, which dictate where container images are pulled from during deployment operations.

The operational impact of this vulnerability is severe due to the privileged nature of the Kubeshark components within a Kubernetes cluster. The deployed worker DaemonSet operates with host network access, host filesystem mounts, and elevated privileges necessary for deep packet inspection via eBPF. When an attacker successfully exploits this command injection flaw, they can redirect these critical infrastructure images—specifically the worker, Hub, and front-end containers—to a registry controlled by the adversary. Since the cluster already possesses permissions to deploy these components, pulling a malicious image from a rogue registry allows for arbitrary code execution with node-level privileges. This effectively grants the attacker full control over the underlying host nodes and potentially escalates to cluster-admin access if RBAC policies are not strictly isolated or if default service accounts have excessive permissions. The ability to hijack container images in this manner bypasses standard security controls that rely on image integrity verification, as the vulnerability occurs at the command construction phase before any signature validation can take place.

From a classification perspective, this issue aligns with CWE-78 Improper Neutralization of Special Elements used in an OS Command and CWE-94 Improper Control of Generation of Code. The attack pattern is consistent with ATT&CK technique T1059 Command and Scripting Interpreter, specifically involving argument injection to manipulate tool behavior. Furthermore, the exploitation path leverages supply chain compromise principles akin to T1608 Install Remote Tool or T1137 Application Desktop Trust Services by hijacking trusted image sources. The root cause is a failure in input sanitization where developer assumptions about string boundaries were violated by allowing special characters that have semantic meaning in command-line parsing contexts.

Mitigation strategies must prioritize immediate patching to version 53.4.0 or later, which addresses the argument handling logic by correctly implementing -- end-of-options separators and validating inputs against expected patterns. In environments where upgrading is not immediately feasible, network-level controls should be enforced to restrict outbound connections from Kubeshark pods to only known, trusted container registries using egress firewall rules or Kubernetes NetworkPolicies. Additionally, enforcing strict RBAC policies that limit the permissions of service accounts associated with Kubeshark components can reduce the blast radius if an image pull is hijacked. Implementing admission controllers such as OPA Gatekeeper or Kyverno to validate pod specifications and prevent privileged containers from running without explicit justification adds another layer of defense. Regular auditing of MCP tool usage and monitoring for anomalous outbound network connections from eBPF-enabled pods are also recommended operational security practices to detect potential exploitation attempts early.

Responsible

GitHub M

Reservation

11/13/2026

Disclosure

11/16/2026

Moderation

accepted

CPE

ready

EPSS

0.00000

KEV

no

Activities

very low

Sources

Do you know our Splunk app?

Download it now for free!