designlang 11.0.1 → 11.0.2

package/bin/design-extract.js

@@ -84,6 +84,7 @@ program
   .option('--user-agent <ua>', 'override the browser User-Agent string')
   .option('--insecure', 'ignore HTTPS/SSL certificate errors (self-signed, dev, proxies)')
   .option('--ignore <selectors...>', 'CSS selectors to remove before extraction')
+  .option('--ignore-widgets', 'Also ignore a curated list of third-party widgets (Intercom, Drift, HubSpot chat, cookie banners, reCAPTCHA, etc.)  See `designlang widgets`.')
   .option('--selector <css>', 'only extract design from elements matching this CSS selector (e.g. ".pricing-card")')
   .option('--system-chrome', 'use the system Chrome install instead of the bundled Chromium (skips the 150MB Playwright download)')
   .option('--tokens-legacy', 'Emit pre-v7 flat token JSON (backward compat)')
@@ -106,6 +107,11 @@ program
     const config = loadConfig();
     const merged = mergeConfig(opts, config);
 
+    if (merged.ignoreWidgets || opts.ignoreWidgets) {
+      const { widgetIgnoreList } = await import('../src/widgets.js');
+      merged.ignore = [...(merged.ignore || []), ...widgetIgnoreList()];
+    }
+
     // Validate URL
     validateUrl(url);
 
@@ -1056,6 +1062,15 @@ program
     }
   });
 
+// ── Widgets — print the curated third-party widget ignore list ─
+program
+  .command('widgets')
+  .description('Print the curated widget-ignore selector list used by --ignore-widgets')
+  .action(async () => {
+    const { WIDGET_SELECTORS } = await import('../src/widgets.js');
+    for (const s of WIDGET_SELECTORS) console.log(s);
+  });
+
 // ── CI command — single PR-comment-ready report ────────────
 program
   .command('ci <url>')

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "designlang",
-  "version": "11.0.1",
+  "version": "11.0.2",
   "description": "Extract the complete design language from any website and ship it — clone to a working Next.js starter, guard tokens with a CI drift bot, or browse everything in a local studio. Outputs W3C DTCG tokens, motion tokens, typed anatomy stubs, Tailwind config, and ready-to-paste v0 / Lovable / Cursor / Claude-Artifacts prompts.",
   "type": "module",
   "bin": {

package/src/widgets.js

@@ -0,0 +1,48 @@
+// Curated selector list for third-party widgets that pollute extractions.
+// Enabled via --ignore-widgets; appended to user-supplied --ignore selectors.
+//
+// Criteria for inclusion: widget renders on tens of thousands of sites, uses a
+// stable ID/attr hook, and its styles would skew token extraction (chat bubbles
+// with off-brand radii/shadows, cookie banners with alarm colors, etc.).
+
+export const WIDGET_SELECTORS = [
+  // Live chat / support
+  '#intercom-container', '#intercom-frame', 'iframe[name*="intercom"]',
+  '#drift-widget', '#drift-frame-controller', '#drift-frame-chat',
+  '#hubspot-messages-iframe-container',
+  '#crisp-chatbox', '.crisp-client',
+  'iframe[title*="Messaging"]', '#launcher', '[data-product="web_widget"]', // Zendesk
+  '#tawk-default', 'iframe[title*="chat window"]',
+  '#chat-widget-container', '#livechat-compact-container', // LiveChat
+  '#helpshift-iframe',
+  'iframe[src*="freshchat"]', '#fc_frame',
+  'iframe[src*="olark"]', '#olark-wrapper',
+
+  // Cookie / consent banners
+  '#CybotCookiebotDialog', '#CybotCookiebotDialogBody',
+  '#onetrust-banner-sdk', '#onetrust-consent-sdk', '#onetrust-pc-sdk',
+  '.termly-banner-top', '.termly-styles-banner',
+  '#cookiebanner', '#cookie-banner', '#cookie-notice',
+  '.cc-window', '.cc-banner', // Cookieconsent by Insites
+  '#usercentrics-root',
+  '#iubenda-cs-banner',
+
+  // reCAPTCHA / anti-bot (visible v2 badge)
+  '.grecaptcha-badge', 'iframe[src*="recaptcha"]',
+
+  // Analytics / pixel iframes (invisible but can leak tokens)
+  'iframe[src*="doubleclick"]', 'iframe[src*="googletagmanager"]',
+  'iframe[src*="facebook.com/tr"]',
+
+  // Social share floating bars
+  '.addthis_floating_style', '#addthis-smartlayers',
+  '.sharethis-inline-share-buttons',
+
+  // Generic catch-alls (last so specific IDs take priority in logs)
+  '[id^="chat-widget"]', '[class*="chat-widget"]',
+  '[aria-label*="cookie" i][role="dialog"]',
+];
+
+export function widgetIgnoreList() {
+  return [...WIDGET_SELECTORS];
+}