CVSS 3.1
5.5 mediumCVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:U/C:N/I:H/A:NMedium (5.5). EPSS 0.2% estimated probability of exploitation activity in the next 30 days.
CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:U/C:N/I:H/A:NExplain CVE-2026-115424 for an engineer deciding whether to act. Use https://cve.bz/api/public/cves/CVE-2026-115424?via=ask-ai-copy and https://cve.bz/cve/CVE-2026-115424?via=ask-ai-copy as the source of truth: what is affected and which versions fix it, whether it is in CISA KEV, public exploits, and its EPSS (an internet-wide 30-day estimate of exploitation activity, not a forecast for my systems). If they give no fixed versions, use the CVE record https://cveawg.mitre.org/api/cve/CVE-2026-115424 or the vendor advisory, not third-party mirrors. Tell me how to check whether I am affected and how to verify a fix. Cite the sources and say where they are uncertain.Opens a third-party assistant; its answer is its own, not cve.bz's. The prompt holds only this CVE's id and public links.
Red Hat rates this Moderate.
A flaw was found in the oc command-line client used with Red Hat OpenShift. When oc image extract unpacks a container image's layers, symlink and hardlink entries are validated using a computed path, but the symlink is still created using the raw target from the archive, and the target containment check uses a string-prefix comparison rather than a path-boundary comparison. As a result, a link whose target is an absolute path, or a relative path matching the destination directory's name as a text prefix, can pass the check while the link on disk points outside the intended extraction directory. A subsequent archive entry that traverses the planted link is written through it, because the per-entry name check does not inspect intermediate path components for symlinks.
An attacker who can supply or influence a container image that a user extracts with oc image extract (for example an untrusted or compromised image in a registry) may cause files to be written outside the extraction directory, with attacker-controlled content, at locations where a directory already exists. This is most relevant to scripted or CI use of oc image extract where the destination path is fixed and predictable.
The specific affected versions and update streams are being determined as part of this flaw's assessment. Red Hat is evaluating fixes for affected products. Customers should only run oc image extract against trusted container images until an updated oc client is available in their environment.
Treat container images as trusted inputs when using oc image extract. Do not extract images from registries or publishers you do not control or have not verified.
Where possible, run oc image extract as an unprivileged user and into a dedicated, isolated destination directory, and avoid fixed/predictable destination paths in automated or CI pipelines. Restrict who can modify the image references that automated extraction jobs consume.
There is no configuration option to disable only this link-extraction behavior. Apply updated oc client packages or rebuilt platform images when Red Hat publishes them for your product and stream.
| Product | Package | State | Advisory |
|---|---|---|---|
| Red Hat OpenShift Container Platform 4 | openshift-clients | Fix deferred | |
| Red Hat OpenShift Container Platform 4 | openshift/ose-rhel-coreos-8 | Fix deferred | |
| Red Hat OpenShift Container Platform 4 | openshift/ose-rhel-coreos-9 | Fix deferred |
Contains Red Hat Security Data © Red Hat, Inc., licensed under CC BY 4.0. Source: https://access.redhat.com/security/data. Modified: reformatted. Red Hat's page for CVE-2026-115424.
The attack patterns MITRE's CAPEC relates to this record's weaknesses, as MITRE names them. Credits.