Vaadin vulnerabilities

A prototype pollution vulnerability exists in the deep merge helpers of Vaadin Charts and Vaadin Component Base. Merging an object the application does not control into a chart configuration or into a component's i18n property writes onto Object.prototype, making the injected properties visible to every object in the running application.

Action captions in Vaadin accept HTML by default but were not sanitized, potentially allowing Cross-site Scripting (XSS) if caption content is derived from user input. In Vaadin Framework 7 and 8, the Action class is general-purpose, while in Vaadin 23 and newer it is only used by the Spreadsheet component. Fixed versions sanitize captions by default and provide an API to explicitly enable HTML content mode.

In the Vaadin Upload component, when the start listener is used to validate metadata about an incoming upload, it is possible to bypass the upload validation. Affected versions include Vaadin 7.0.0-7.7.47, 8.0.0-8.28.1, 14.0.0-14.13.0, 23.0.0-23.6.1, and 24.0.0-24.7.6.

In Vaadin, in many versions, a vulnerability allows information disclosure of class and method names in RPC responses by sending modified requests.

In Vaadin, adding non-visible components to the UI on the server side causes their content to be sent to the browser, potentially leading to information disclosure. Affected versions include 10.0.0-10.0.22, 11.0.0-14.10.0, 15.0.0-22.0.28, 23.0.0-23.3.12, 24.0.0-24.0.5, and 24.1.0.alpha1-24.1.0.beta1.

The default configuration of the TreeGrid component uses Object::toString as a key in client-side and server communication in Vaadin, potentially disclosing values that should not be available on the client side. Affected versions include Vaadin 14.8.5-14.8.9, 22.0.6-22.0.14, 23.0.0.beta2-23.0.8, and 23.1.0.alpha1-23.1.0.alpha4.