npm · mathlive
mathlive's Lack of Escaping of HTML allows for XSS
Despite the 0.104.0 patch escaping attribute-bearing constructs (\htmlData, \href), text-content reflection was missed. The \text{}, \mbox{} commands accept arbitrary characters in their body and emit them raw and unescaped into both the HTML markup and the MathML output, leading to XSS.
Box.toMarkup at src/core/box.ts:356 concatenates this.value into the rendered span without HTML-escaping. In text mode any literal character (<, >, &, ") is wrapped into a TextAtom whose value is the raw codepoint and lands in the markup unchanged. The MathML serializer at src/formats/atom-to-math-ml.ts is independently broken: xmlEscape deliberately omits the & rule, and scanText, case 'text', and the mode === 'text' early return all emit atom.value raw.
Both outputs flow into innerHTML sinks via the public API. <math-span> / <math-div> (src/public/math-static-elements.ts:331,407) bypass MathfieldElement.createHTML entirely. The editor and SSR paths route through createHTML, but its default value is the identity function (src/public/mathfield-element.ts:789).
const s = document.createElement('math-span');
s.style.display = 'block';
s.textContent = '\\text{<img src=x onerror=alert(1)>}';
document.body.appendChild(s);
s.scrollIntoView();
Equivalent payloads: \mbox{<img src=x onerror=alert(1)>}
or
import { convertLatexToMarkup } from 'mathlive';
document.body.innerHTML = convertLatexToMarkup('\\text{<img src=x onerror=alert(1)>}');
MathLive users who render untrusted mathematical expressions can encounter malicious input that runs arbitrary JavaScript.
Is your project exposed to this? Stateward checks every dependency on every pull request and flags it only if your code actually reaches it.
Check my repoSources: CISA KEV (public domain), OSV.dev & GitHub Advisory Database (CC-BY-4.0), FIRST EPSS, NVD/CWE (public domain). Served live from the Stateward advisory database.