CVE-2026-84664 in GitLab Plugin
Summary
by MITRE • 09/02/2026
Jenkins GitLab Plugin 1.9.16 and earlier allows overwriting the global GitLab connection configuration through Stapler data binding, allowing attackers to connect to an attacker-specified URL using GitLab API tokens already configured by administrators.
If you want to get the best quality for vulnerability data then you always have to consider VulDB.
Analysis
by VulDB Data Team • 09/02/2026
The Jenkins GitLab plugin version 1.9.16 and prior versions contain a critical security flaw rooted in improper validation of user-supplied input during the configuration update process. This vulnerability arises from the use of Stapler, a web framework component within Jenkins that handles data binding between HTTP requests and Java objects. Specifically, when an authenticated administrator attempts to modify or save global GitLab connection settings, the application relies on automatic property mapping without sufficient sanitization or validation of the input fields. An attacker who has gained access to the Jenkins interface with administrative privileges can exploit this mechanism by crafting a malicious request that manipulates these bound properties. The core technical issue is a lack of strict type checking and domain validation for the server URL field, allowing an adversary to inject arbitrary endpoints rather than being restricted to legitimate GitLab instances.
This flaw effectively allows privilege escalation in terms of data integrity and potential credential misuse within the CI/CD pipeline context. By overwriting the global configuration, the attacker redirects all subsequent API calls made by Jenkins to a malicious server controlled by them. Since Jenkins typically stores sensitive authentication credentials, such as GitLab personal access tokens or OAuth secrets, alongside these connection settings, the compromised endpoint can intercept and exfiltrate these high-value assets. The operational impact is severe because it undermines the trust boundary between the CI/CD system and its source code management platform. Once an attacker possesses valid API tokens for a major version control provider like GitLab, they can clone private repositories, inject malicious code into branches, or delete critical project history, leading to supply chain compromise and significant intellectual property loss.
From a classification perspective, this vulnerability aligns with CWE-20 Improper Input Validation, as the system fails to verify that the provided URL conforms to expected patterns or belongs to an authorized domain. It also relates to CWE-918 Server-Side Request Forgery (SSRF) principles insofar as it forces the server to make requests to a location determined by external input, albeit with administrative context rather than anonymous access. In terms of the MITRE ATT&CK framework for enterprise security, this behavior maps to T1078 Valid Accounts, where an attacker uses legitimate credentials to gain persistence and control over system configurations. Furthermore, it touches upon T1562 Impair Defenses by altering security-related settings that protect data integrity within the development environment.
Mitigation strategies must prioritize immediate patching of the Jenkins GitLab plugin to version 1.9.17 or later, where this binding issue has been addressed with stricter input validation and domain allowlisting mechanisms. Administrators should also enforce strict access controls on Jenkins configuration pages, ensuring that only trusted individuals can modify global connection settings. Implementing a Web Application Firewall (WAF) ruleset to detect anomalous POST requests targeting Stapler endpoints may provide an additional layer of defense during the transition period. Additionally, organizations should rotate any GitLab API tokens associated with compromised or potentially exposed Jenkins instances immediately, as past exposure cannot be fully ruled out given the nature of configuration overwrites. Regular auditing of plugin versions and continuous monitoring for unauthorized changes to global configurations are essential practices to maintain the integrity of automated development pipelines.