diff --git a/scripts/lint-404s/ignore-list.txt b/scripts/lint-404s/ignore-list.txt
new file mode 100644
index 0000000000000..17a5d5eae18e3
--- /dev/null
+++ b/scripts/lint-404s/ignore-list.txt
@@ -0,0 +1,2 @@
+/
+/changelog/
diff --git a/scripts/lint-404s/main.ts b/scripts/lint-404s/main.ts
new file mode 100644
index 0000000000000..8bcbfde329aaa
--- /dev/null
+++ b/scripts/lint-404s/main.ts
@@ -0,0 +1,149 @@
+/* eslint-disable no-console */
+
+import {readFileSync} from 'fs';
+import path, {dirname} from 'path';
+import {fileURLToPath} from 'url';
+
+const baseURL = 'http://localhost:3000/';
+type Link = {href: string; innerText: string};
+
+const trimSlashes = (s: string) => s.replace(/(^\/|\/$)/g, '');
+
+// @ts-ignore
+const ignoreListFile = path.join(dirname(import.meta.url), './ignore-list.txt');
+
+// Paths to skip
+const ignoreList: string[] = readFileSync(fileURLToPath(ignoreListFile), 'utf8')
+  .split('\n')
+  .map(trimSlashes)
+  .filter(Boolean);
+
+async function fetchWithFollow(url: URL | string): Promise<Response> {
+  const r = await fetch(url);
+  if (r.status >= 300 && r.status < 400 && r.headers.has('location')) {
+    return fetchWithFollow(r.headers.get('location')!);
+  }
+  return r;
+}
+
+async function main() {
+  const sitemap = await fetch(`${baseURL}sitemap.xml`).then(r => r.text());
+
+  const slugs = [...sitemap.matchAll(/<loc>([^<]*)<\/loc>/g)]
+    .map(l => l[1])
+    .map(url => trimSlashes(new URL(url).pathname))
+    .filter(Boolean);
+  const allSlugsSet = new Set(slugs);
+
+  console.log('Checking 404s on %d pages', slugs.length);
+
+  const all404s: {page404s: Link[]; slug: string}[] = [];
+
+  // check if the slug equivalent of the href is in the sitemap
+  const isInSitemap = (href: string) => {
+    // remove hash
+    const pathnameSlug = trimSlashes(href.replace(/#.*$/, ''));
+
+    // some #hash links result in empty slugs when stripped
+    return pathnameSlug === '' || allSlugsSet.has(pathnameSlug);
+  };
+
+  function shoudlSkipLink(href: string) {
+    const isExternal = (href_: string) =>
+      href_.startsWith('http') || href_.startsWith('mailto:');
+    const isLocalhost = (href_: string) =>
+      href_.startsWith('http') && new URL(href_).hostname === 'localhost';
+    const isIp = (href_: string) => /(\d{1,3}\.){3}\d{1,3}/.test(href_);
+    const isImage = (href_: string) => /\.(png|jpg|jpeg|gif|svg|webp)$/.test(href_);
+
+    return [
+      isExternal,
+      (s = '') => ignoreList.includes(trimSlashes(s)),
+      isImage,
+      isLocalhost,
+      isIp,
+    ].some(fn => fn(href));
+  }
+
+  async function is404(link: Link, pageUrl: URL): Promise<boolean> {
+    if (shoudlSkipLink(link.href)) {
+      return false;
+    }
+
+    const fullPath = link.href.startsWith('/')
+      ? trimSlashes(link.href)
+      : // relative path
+        trimSlashes(new URL(pageUrl.pathname + '/' + link.href, baseURL).pathname);
+
+    if (isInSitemap(fullPath)) {
+      return false;
+    }
+    const fullUrl = new URL(fullPath, baseURL);
+    const resp = await fetchWithFollow(fullUrl);
+    if (resp.status === 404) {
+      return true;
+    }
+    return false;
+  }
+
+  for (const slug of slugs) {
+    const pageUrl = new URL(slug, baseURL);
+    const now = performance.now();
+    const html = await fetchWithFollow(pageUrl.href).then(r => r.text());
+
+    const linkRegex = /<a[^>]*href="([^"]*)"[^>]*>([^<]*)<\/a>/g;
+    const links = Array.from(html.matchAll(linkRegex)).map(m => {
+      const [, href, innerText] = m;
+      return {href, innerText};
+    });
+    const page404s = (
+      await Promise.all(
+        links.map(async link => {
+          const is404_ = await is404(link, pageUrl);
+          return [link, is404_] as [Link, boolean];
+        })
+      )
+    )
+      .filter(([_, is404_]) => is404_)
+      .map(([link]) => link);
+
+    if (page404s.length) {
+      all404s.push({slug, page404s});
+    }
+
+    console.log(
+      page404s.length ? '❌' : '✅',
+      `in ${(performance.now() - now).toFixed(1).padStart(4, '0')} ms | ${slug}`
+    );
+  }
+
+  if (all404s.length === 0) {
+    console.log('\n\n🎉 No 404s found');
+    return false;
+  }
+  console.log(
+    '\n❌ Found %d 404s across %d %s',
+    all404s.map(x => x.page404s.length).reduce((a, b) => a + b, 0),
+    all404s.length,
+    all404s.length === 1 ? 'page' : 'pages'
+  );
+  for (const {slug, page404s} of all404s) {
+    console.log('\n🌐', baseURL + slug);
+    for (const link of page404s) {
+      console.log(`    - [${link.innerText}](${link.href})`);
+    }
+  }
+
+  console.log(
+    '\n👉 Note: the markdown syntax is not necessarily on the source files, but the links do exist on the final pages'
+  );
+  // signal error
+  return true;
+}
+const now = performance.now();
+main().then(has404s => {
+  console.log(`\n Done in ${(performance.now() - now).toFixed(1)} ms`);
+  process.exit(has404s ? 1 : 0);
+});
+
+export {};