security: block data:image/svg+xml + sandbox HtmlBlock iframes

M-5: safeUrl() blocked javascript:/vbscript:/data:text/html but
allowed data:image/svg+xml, which can execute inline <script>/onload=
when loaded as a document/navigation target despite its "image" MIME
type (defense in depth -- not currently reachable to execution via
this sink, but closing it). Added `data:image/svg+xml` to the existing
DANGEROUS_SCHEME_PREFIXES check, so it's caught after the same
entity-decode/whitespace-strip/lowercase normalization used for the
other blocked schemes (obfuscated variants included). Other
data:image/* types (png/jpeg/gif/webp, ...) remain allowed unchanged.

M-6: HtmlBlock's purifyHtml() allowed <iframe src> through with no
`sandbox` attribute -- a clickjacking/phishing vector even with
DOMPurify already stripping script/on*=. Added a DOMPurify
afterSanitizeAttributes hook, scoped tightly to each purifyHtml() call
(added right before sanitize(), removed in a finally right after) so
it can't leak onto other DOMPurify uses or accumulate duplicates
across repeated calls, that force-sets a restrictive sandbox
(allow-scripts allow-same-origin allow-popups allow-forms -- no
allow-top-navigation) and referrerpolicy=no-referrer on every iframe
that survives sanitization.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This commit is contained in:
2026-07-12 18:31:25 -07:00
parent 86455413d0
commit bf4a9f48eb
4 changed files with 98 additions and 2 deletions
+8 -1
View File
@@ -36,7 +36,14 @@ export function escapeAttr(s: string): string {
// Dangerous scheme prefixes, checked against a normalized copy with all
// colons removed (see safeUrl) so that colon-splicing obfuscation like
// "java:script:alert(1)" can't slip past a literal "javascript:" check.
const DANGEROUS_SCHEME_PREFIXES = ['javascript', 'vbscript', 'datatext/html'];
// M-5: `data:image/svg+xml` is blocked alongside `data:text/html` -- an SVG
// document can carry an inline <script>/onload= just like an HTML document,
// so it executes script when loaded as a document/navigation target (e.g. an
// <a href> or window.open), even though it's nominally an "image" MIME type.
// Other `data:image/*` types (png/jpeg/gif/webp, ...) stay allowed below --
// only this specific scriptable subtype is blocked, regardless of a
// trailing `;base64` or other parameters.
const DANGEROUS_SCHEME_PREFIXES = ['javascript', 'vbscript', 'datatext/html', 'dataimage/svg+xml'];
/** Decodes &#NN; and &#xNN; numeric HTML entities (used to obfuscate scheme names). */
function decodeNumericEntities(s: string): string {