fix: sanitize data-driven href URLs

Co-authored-by: Sebastien Tardif <sebtardif@ncf.ca>

Author: steipete

CHANGELOG.md

@@ -3,6 +3,7 @@
 ## Unreleased
 
 - Ecosystem: promote clawpdf into the TypeScript libraries section with its canonical site link.
+- Website: sanitize data-driven external links before rendering them into `href` attributes (#143, thanks @SebTardif).
 - Website: move the Discord shortcut into Vercel routing so `/discord` redirects correctly (#147, thanks @SebTardif).
 - Ecosystem: tune project card banner art visibility.
 - Ecosystem: quiet the final contribution CTA button colors.

package.json

@@ -5,6 +5,7 @@
   "scripts": {
     "dev": "astro dev",
     "build": "astro build",
+    "test": "bun test",
     "avatars:cache": "node scripts/cache-avatars.mjs",
     "preview": "astro preview"
   },

src/lib/sanitize-url.ts

@@ -0,0 +1,21 @@
+const allowedProtocols = new Set(['http:', 'https:', 'mailto:']);
+
+function normalizeForProtocolCheck(url: string): string {
+  let decoded = url;
+  try {
+    decoded = decodeURIComponent(url);
+  } catch {
+    // Malformed escapes still flow through URL parsing below.
+  }
+
+  return decoded.replace(/[\x00-\x1f\x7f]/g, '');
+}
+
+export function sanitizeUrl(url: string): string {
+  try {
+    const parsed = new URL(normalizeForProtocolCheck(url), 'https://openclaw.ai');
+    return allowedProtocols.has(parsed.protocol) ? url : '#';
+  } catch {
+    return '#';
+  }
+}

src/pages/blog/[...slug].astro

@@ -4,6 +4,7 @@ import SiteTopbar from '../../components/SiteTopbar.astro';
 import { render } from 'astro:content';
 import { getPublishedBlogPosts } from '../../lib/blog';
 import { getCachedXAvatarSrc, getInitialsAvatarSrc } from '../../lib/avatars';
+import { sanitizeUrl } from '../../lib/sanitize-url';
 
 export async function getStaticPaths() {
   const posts = await getPublishedBlogPosts();
@@ -127,7 +128,7 @@ const postUrl = `https://openclaw.ai/blog/${post.id}`;
                   {getAuthorLinks(author).length > 0 && (
                     <div class="author-links">
                       {getAuthorLinks(author).map((link) => (
-                        <a href={link.url} class="author-handle" target="_blank" rel="noopener">{link.label}</a>
+                        <a href={sanitizeUrl(link.url)} class="author-handle" target="_blank" rel="noopener">{link.label}</a>
                       ))}
                     </div>
                   )}

src/pages/index.astro

@@ -8,6 +8,7 @@ import communityBuilds from '../data/community-builds.json';
 import pressArticles from '../data/press.json';
 import { getPublishedBlogPosts } from '../lib/blog';
 import { getCachedXAvatarSrc, getInitialsAvatarSrc } from '../lib/avatars';
+import { sanitizeUrl } from '../lib/sanitize-url';
 
 const blogPosts = await getPublishedBlogPosts();
 const [latestPost] = blogPosts;
@@ -534,7 +535,7 @@ function formatBlogDate(date: Date): string {
       <div class="testimonials-track">
         <div class="testimonials-row row-1" style={`--duration: ${duration1}s`}>
           {row1.map((t) => (
-            <a href={t.url} target="_blank" rel="noopener" class="testimonial-card" tabindex="-1">
+            <a href={sanitizeUrl(t.url)} target="_blank" rel="noopener" class="testimonial-card" tabindex="-1">
               <img
                 src={getCachedXAvatarSrc(t.author, t.author, 88)}
                 alt={t.author}
@@ -551,7 +552,7 @@ function formatBlogDate(date: Date): string {
         </div>
         <div class="testimonials-row row-2" style={`--duration: ${duration2}s`}>
           {row2.map((t) => (
-            <a href={t.url} target="_blank" rel="noopener" class="testimonial-card">
+            <a href={sanitizeUrl(t.url)} target="_blank" rel="noopener" class="testimonial-card">
               <img
                 src={getCachedXAvatarSrc(t.author, t.author, 88)}
                 alt={t.author}
@@ -666,7 +667,7 @@ function formatBlogDate(date: Date): string {
       />
       <div class="builds-grid">
         {featuredBuilds.map((build) => (
-          <a href={build.href} target="_blank" rel="noopener" class="build-card">
+          <a href={sanitizeUrl(build.href)} target="_blank" rel="noopener" class="build-card">
             <div class="build-card-header">
               <span class="build-source">{build.source}</span>
               <span class="build-open">Open →</span>
@@ -694,7 +695,7 @@ function formatBlogDate(date: Date): string {
       <div class="press-grid">
         {featuredPress.map((article) => (
           <a
-            href={article.url}
+            href={sanitizeUrl(article.url)}
             target="_blank"
             rel="noopener"
             class:list={['press-card', article.featured && 'press-featured']}

src/pages/press.astro

@@ -1,6 +1,7 @@
 ---
 import Layout from '../layouts/Layout.astro';
 import pressArticles from '../data/press.json';
+import { sanitizeUrl } from '../lib/sanitize-url';
 ---
 
 <Layout
@@ -26,7 +27,7 @@ import pressArticles from '../data/press.json';
     <div class="press-grid">
       {pressArticles.map((article) => (
         <a
-          href={article.url}
+          href={sanitizeUrl(article.url)}
           target="_blank"
           rel="noopener"
           class:list={['press-card', article.featured && 'press-featured']}

src/pages/shoutouts.astro

@@ -3,6 +3,7 @@ import Layout from '../layouts/Layout.astro';
 import testimonials from '../data/testimonials.json';
 import extraTestimonials from '../data/testimonials-extra.json';
 import { getCachedXAvatarSrc, getInitialsAvatarSrc } from '../lib/avatars';
+import { sanitizeUrl } from '../lib/sanitize-url';
 
 // Combine all testimonials - strongest first, then extras (no randomization)
 const allTestimonials = [...testimonials, ...extraTestimonials];
@@ -23,7 +24,7 @@ const allTestimonials = [...testimonials, ...extraTestimonials];
 
     <div class="shoutouts-grid">
       {allTestimonials.map((t) => (
-        <a href={t.url} target="_blank" rel="noopener" class="shoutout-card">
+        <a href={sanitizeUrl(t.url)} target="_blank" rel="noopener" class="shoutout-card">
           <img
             src={t.avatar || getCachedXAvatarSrc(t.author, t.author, 96)}
             alt={t.author}

src/pages/showcase.astro

@@ -3,6 +3,7 @@ import Layout from '../layouts/Layout.astro';
 import showcaseData from '../data/showcase.json';
 import featuredBuilds from '../data/community-builds.json';
 import { getCachedXAvatarSrc, getInitialsAvatarSrc } from '../lib/avatars';
+import { sanitizeUrl } from '../lib/sanitize-url';
 
 const featuredIds = new Set(featuredBuilds.map((item) => item.id));
 const sortedShowcase = showcaseData.filter((item) => !featuredIds.has(item.id));
@@ -41,7 +42,7 @@ const categoryLabels: Record<string, string> = {
       </div>
       <div class="featured-grid">
         {featuredBuilds.map((build) => (
-          <a href={build.href} target="_blank" rel="noopener" class="featured-card">
+          <a href={sanitizeUrl(build.href)} target="_blank" rel="noopener" class="featured-card">
             <div class="featured-topline">
               <span class="featured-source">{build.source}</span>
               <span class="featured-likes">❤️ {build.likes}</span>

tests/sanitize-url.test.ts

@@ -0,0 +1,30 @@
+import { describe, expect, test } from 'bun:test';
+import { sanitizeUrl } from '../src/lib/sanitize-url';
+
+describe('sanitizeUrl', () => {
+  test('keeps allowed absolute and relative URLs unchanged', () => {
+    expect(sanitizeUrl('https://x.com/openclaw')).toBe('https://x.com/openclaw');
+    expect(sanitizeUrl('http://example.com/path')).toBe('http://example.com/path');
+    expect(sanitizeUrl('mailto:security@openclaw.ai')).toBe('mailto:security@openclaw.ai');
+    expect(sanitizeUrl('/blog#security')).toBe('/blog#security');
+  });
+
+  test('blocks dangerous protocols', () => {
+    expect(sanitizeUrl('javascript:alert(1)')).toBe('#');
+    expect(sanitizeUrl('JaVaScRiPt:alert(1)')).toBe('#');
+    expect(sanitizeUrl('data:text/html,<script>alert(1)</script>')).toBe('#');
+    expect(sanitizeUrl('vbscript:msgbox(1)')).toBe('#');
+    expect(sanitizeUrl('ftp://example.com/file')).toBe('#');
+  });
+
+  test('blocks encoded and control-character protocol bypasses', () => {
+    expect(sanitizeUrl('%6a%61%76%61%73%63%72%69%70%74:alert(1)')).toBe('#');
+    expect(sanitizeUrl('java\u0000script:alert(1)')).toBe('#');
+    expect(sanitizeUrl('java\nscript:alert(1)')).toBe('#');
+  });
+
+  test('blocks unparsable input', () => {
+    expect(sanitizeUrl('http://[::1')).toBe('#');
+    expect(sanitizeUrl('javascript:%E0%A4%A')).toBe('#');
+  });
+});