Disclosed vulnerabilities
Fixed: Oct 15, 2025
CVE-2025-34281 Published: Nov 27, 2025

Stored Cross-Site Scripting (XSS) in ThingsBoard

6.2 Medium Severity IOT JAVA Stored-XSS XSS
João Oliveira Reported by Checkmarx Zero

Summary

ThingsBoard in versions prior to v4.2.1 allows an authenticated user to upload malicious SVG images via the “Image Gallery”, leading to a Stored Cross-Site Scripting (XSS) vulnerability. The exploit can be triggered when any user accesses the public API endpoint of the malicious SVG images, or if the malicious images are embedded in an iframe element, during a widget creation, deployed to any page of the platform (e.g., dashboards), and accessed during normal operations. The vulnerability resides in the ImageController, which fails to restrict the execution of JavaScript code when an image is loaded by the user’s browser. This vulnerability can lead to the execution of malicious code in the context of other users’ sessions, potentially compromising their accounts and allowing unauthorized actions.

Product

This vulnerability affects ThingsBoard versions prior to v4.2.1.

Impact

This vulnerability can lead to the execution of malicious code in the context of other users’ sessions, potentially compromising their accounts and allowing unauthorized actions.

Steps to reproduce

  1. Craft a simple payload to be injected into an SVG image that pops up a JavaScript alert:

    <svg xmlns="http://www.w3.org/2000/svg" width="200" height="200" viewBox="0 0 200 200">
     <!-- Visual SVG -->
     <rect width="200" height="200" fill="#3498db" />
    
     <!-- Text in the center -->
     <text x="50%" y="50%" font-size="20" text-anchor="middle" dy=".3em" fill="white">Proof of Concept</text>
    
     <!-- XSS Payload: Pop up an alert box with the JWT -->
     <script type="text/javascript">
         <![CDATA[
         alert(JSON.stringify(localStorage));
         ]]>
     </script>
    </svg>
  2. Login to ThingsBoard as a tenant administrator

  3. Upload the SVG image using the Image Gallery

  4. Copy the public image API endpoint for the malicious SVG image

  5. Access the public image API endpoint

Expected result

The XSS payload is executed as soon as anyone accesses the malicious SVG image public API endpoint. It is also possible to embed the SVG image in an iframe by creating an widget, and deploying it to any page, like a dashboard. If a user during normal browsing opens the page with the malicious widget, the XSS payload will be executed.

Remediation

Fixed in ThingsBoard v4.2.1 by adding the Content-Security-Policy header with the default-src 'none' value. This prevents external resources (scripts, styles, frames, etc.) being loaded from uploaded files, including SVG images.

Credit

This issue was discovered and reported by Checkmarx Security Researcher João Oliveira.

Resources

  1. NVD entry for CVE-2025-34281
  2. Fix Commit

Stay connected and never miss a research update!