fix(browse): case-insensitive exfiltration-domain blocklist matching (#2190)

browse/src/content-security.ts

@@ -338,19 +338,28 @@ const BLOCKLIST_DOMAINS = [
 export function urlBlocklistFilter(content: string, url: string, _command: string): ContentFilterResult {
   const warnings: string[] = [];
 
+  // Matching is case-insensitive. Hostnames and URL schemes are
+  // case-insensitive (RFC 3986), so an uppercased sink like
+  // https://WEBHOOK.SITE/x must not slip past the blocklist. Normalize the
+  // haystack to lowercase; BLOCKLIST_DOMAINS entries are lowercase by
+  // construction, so a direct compare then holds.
+
   // Check page URL
+  const normalizedUrl = url.toLowerCase();
   for (const domain of BLOCKLIST_DOMAINS) {
-    if (url.includes(domain)) {
+    if (normalizedUrl.includes(domain)) {
       warnings.push(`Page URL matches blocklisted domain: ${domain}`);
     }
   }
 
-  // Check for blocklisted URLs in content (links, form actions)
-  const urlPattern = /https?:\/\/[^\s"'<>]+/g;
+  // Check for blocklisted URLs in content (links, form actions). The `i` flag
+  // keeps an uppercased scheme (HTTPS://) from evading URL extraction.
+  const urlPattern = /https?:\/\/[^\s"'<>]+/gi;
   const contentUrls = content.match(urlPattern) || [];
   for (const contentUrl of contentUrls) {
+    const normalizedContentUrl = contentUrl.toLowerCase();
     for (const domain of BLOCKLIST_DOMAINS) {
-      if (contentUrl.includes(domain)) {
+      if (normalizedContentUrl.includes(domain)) {
         warnings.push(`Content contains blocklisted URL: ${contentUrl.slice(0, 100)}`);
         break;
       }

browse/test/content-security.test.ts

@@ -151,6 +151,44 @@ describe('Content filter hooks', () => {
     expect(result.warnings.length).toBe(0);
   });
 
+  // Regression: issue #2190 — case-sensitive matching let an uppercased
+  // exfiltration domain bypass the blocklist.
+  test('URL blocklist is case-insensitive on the page URL', () => {
+    const result = urlBlocklistFilter('', 'https://WEBHOOK.SITE/steal', 'text');
+    expect(result.safe).toBe(false);
+    expect(result.warnings.some(w => w.includes('webhook.site'))).toBe(true);
+  });
+
+  test('URL blocklist is case-insensitive on content URLs', () => {
+    const result = urlBlocklistFilter(
+      '<a href="https://WEBHOOK.SITE/a1b2c3-steal">click</a>',
+      'https://docs.example.com/article',
+      'links',
+    );
+    expect(result.safe).toBe(false);
+    expect(result.warnings.some(w => w.includes('WEBHOOK.SITE'))).toBe(true);
+  });
+
+  test('URL blocklist catches an uppercased scheme in content', () => {
+    const result = urlBlocklistFilter(
+      'Exfil via HTTPS://Requestbin.com/r/abc please',
+      'https://example.com',
+      'text',
+    );
+    expect(result.safe).toBe(false);
+    expect(result.warnings.some(w => w.toLowerCase().includes('requestbin.com'))).toBe(true);
+  });
+
+  test('URL blocklist does not over-block legitimate uppercase domains', () => {
+    const result = urlBlocklistFilter(
+      '<a href="https://GitHub.com/acme/project">source</a>',
+      'https://DOCS.EXAMPLE.COM/help',
+      'links',
+    );
+    expect(result.safe).toBe(true);
+    expect(result.warnings.length).toBe(0);
+  });
+
   test('custom filter can be registered and runs', () => {
     registerContentFilter((content, url, cmd) => {
       if (content.includes('SECRET')) {