
PEACH
Un cadre d’isolation des locataires
CVE-2026-15792 is a Denial of Service vulnerability in Moby BuildKit, the toolkit for building container images, where a malicious BuildKit client or frontend could craft a request that causes the BuildKit daemon to crash with a panic. It affects all BuildKit versions up to and including v0.31.1, with the fix introduced in v0.31.2. The vulnerability was published on July 21, 2026, and assigned a CVSS v3.1 base score of 7.5 (High) and a CVSS v4.0 base score of 6.0 (Medium) (GitHub Advisory, Feedly).
The root cause is classified as CWE-20 (Improper Input Validation): BuildKit does not adequately validate parameters received from a client or frontend before processing them, allowing a specially crafted request to trigger an unhandled panic in the daemon process. The attack vector is network-based, requiring no user interaction, and in the CVSS v3.1 assessment, no privileges are required; however, the CVSS v4.0 assessment notes that some attack requirements (such as the ability to connect to the BuildKit daemon) must be present. The vulnerable code path is exercised when a malicious or compromised frontend — typically specified via the #syntax directive in a Dockerfile or the --frontend flag with buildctl build — sends incorrect parameters to the daemon (GitHub Advisory).
Successful exploitation results in a Denial of Service: the BuildKit daemon crashes with a panic, interrupting all ongoing and queued container image builds. There is no confidentiality or integrity impact — the vulnerability is limited to availability. In CI/CD environments where BuildKit is a critical build infrastructure component, repeated exploitation could cause sustained service disruption and pipeline failures (GitHub Advisory, Feedly).
There is no public proof-of-concept exploit and no evidence of in-the-wild exploitation as of the time of publication (Feedly). The CVE is not listed in the CISA Known Exploited Vulnerabilities catalog. The EPSS score is approximately 0.38%, indicating a low probability of exploitation in the near term. NVD's SSVC assessment classifies exploitation as "none" and the vulnerability as not automatable, reflecting the requirement for an attacker to have access to the BuildKit daemon (Feedly).
#syntax image in a Dockerfile or using buildctl build --frontend with a crafted frontend.panic: followed by a goroutine dump) around the time of a build request from an unfamiliar frontend image.buildkitd process without a graceful shutdown signal; process restart events in system logs (e.g., systemd reporting buildkitd service restarted).#syntax directives pointing to external or unknown registries) (GitHub Advisory).Upgrade BuildKit to v0.31.2 or later, which contains the fix for this vulnerability (GitHub Advisory). As a workaround, avoid using BuildKit frontends from untrusted sources — only allow frontend images specified via #syntax or --frontend from verified, trusted registries. Additionally, restrict network access to the BuildKit daemon to only trusted clients using firewall rules or network-level access controls, and monitor daemon logs for unexpected crashes or suspicious build requests (Feedly). Red Hat has also issued advisories (RHSA-2026:44152, RHSA-2026:46988) for affected downstream packages (Red Hat Advisory 1, Red Hat Advisory 2).
The advisory was published by BuildKit maintainer tonistiigi on GitHub on July 16, 2026, and assigned a "Moderate" severity rating by the project (GitHub Advisory). Red Hat issued errata advisories (RHSA-2026:44152 and RHSA-2026:46988) addressing the vulnerability in downstream products. Coverage has been limited to automated vulnerability tracking platforms such as Vulners, VulDB, and CVEfeed, with no notable independent researcher commentary or significant social media discussion observed (Feedly).
Source: Ce rapport a été généré à l’aide de l’IA
Évaluation gratuite des vulnérabilités
Évaluez vos pratiques de sécurité cloud dans 9 domaines de sécurité pour évaluer votre niveau de risque et identifier les failles dans vos défenses.
Obtenez une démo personnalisée
"La meilleure expérience utilisateur que j’ai jamais vue, offre une visibilité totale sur les workloads cloud."
"Wiz fournit une interface unique pour voir ce qui se passe dans nos environnements cloud."
"Nous savons que si Wiz identifie quelque chose comme critique, c’est qu’il l’est réellement."