From c5851370bf072644177b3ae1e7a94a66ae2331f5 Mon Sep 17 00:00:00 2001
From: Marc Mintel <marc@mintel.me>
Date: Wed, 25 Feb 2026 00:16:20 +0100
Subject: [PATCH] feat: implement robust full-sitemap HTTP validation in smoke
 test phase

---
 .gitea/workflows/deploy.yml |  5 +++
 package.json                |  1 +
 scripts/check-http.ts       | 74 +++++++++++++++++++++++++++++++++++++
 3 files changed, 80 insertions(+)
 create mode 100644 scripts/check-http.ts

diff --git a/.gitea/workflows/deploy.yml b/.gitea/workflows/deploy.yml
index 848dd28a..b8335de2 100644
--- a/.gitea/workflows/deploy.yml
+++ b/.gitea/workflows/deploy.yml
@@ -396,6 +396,11 @@ jobs:
         env:
           TEST_URL: ${{ needs.prepare.outputs.next_public_url }}
         run: pnpm run check:og
+      - name: 🌐 Run Full Sitemap HTTP Validation
+        env:
+          NEXT_PUBLIC_BASE_URL: ${{ needs.prepare.outputs.next_public_url }}
+          GATEKEEPER_PASSWORD: ${{ secrets.GATEKEEPER_PASSWORD || 'klz2026' }}
+        run: pnpm run check:http
 
   # ──────────────────────────────────────────────────────────────────────────────
   # JOB 6: Lighthouse (Performance & Accessibility)
diff --git a/package.json b/package.json
index 73309110..8389abcb 100644
--- a/package.json
+++ b/package.json
@@ -105,6 +105,7 @@
     "check:a11y": "pa11y-ci",
     "check:wcag": "tsx ./scripts/wcag-sitemap.ts",
     "check:html": "tsx ./scripts/check-html.ts",
+    "check:http": "tsx ./scripts/check-http.ts",
     "check:spell": "cspell \"content/**/*.{md,mdx}\" \"app/**/*.tsx\" \"components/**/*.tsx\"",
     "check:security": "tsx ./scripts/check-security.ts",
     "check:links": "bash ./scripts/check-links.sh",
diff --git a/scripts/check-http.ts b/scripts/check-http.ts
new file mode 100644
index 00000000..c279ee9c
--- /dev/null
+++ b/scripts/check-http.ts
@@ -0,0 +1,74 @@
+import axios from 'axios';
+import * as cheerio from 'cheerio';
+
+const targetUrl = process.argv[2] || process.env.NEXT_PUBLIC_BASE_URL || 'http://localhost:3000';
+const gatekeeperPassword = process.env.GATEKEEPER_PASSWORD || 'klz2026';
+
+async function main() {
+  console.log(`\n🚀 Starting HTTP Sitemap Validation for: ${targetUrl}\n`);
+
+  try {
+    const sitemapUrl = `${targetUrl.replace(/\/$/, '')}/sitemap.xml`;
+    console.log(`📥 Fetching sitemap from ${sitemapUrl}...`);
+
+    const response = await axios.get(sitemapUrl, {
+      headers: { Cookie: `klz_gatekeeper_session=${gatekeeperPassword}` },
+      validateStatus: (status) => status < 400,
+    });
+
+    const $ = cheerio.load(response.data, { xmlMode: true });
+    let urls = $('url loc')
+      .map((i, el) => $(el).text())
+      .get();
+
+    const urlPattern = /https?:\/\/[^\/]+/;
+    urls = [...new Set(urls)]
+      .filter((u) => u.startsWith('http'))
+      .map((u) => u.replace(urlPattern, targetUrl.replace(/\/$/, '')))
+      .sort();
+
+    console.log(`✅ Found ${urls.length} target URLs in sitemap.`);
+
+    if (urls.length === 0) {
+      console.error('❌ No URLs found in sitemap. Is the site up?');
+      process.exit(1);
+    }
+
+    console.log(`\n🔍 Verifying HTTP Status Codes (Limit: None)...`);
+    let hasErrors = false;
+
+    // Run fetches sequentially to avoid overwhelming the server during CI
+    for (let i = 0; i < urls.length; i++) {
+      const u = urls[i];
+      try {
+        const res = await axios.get(u, {
+          headers: { Cookie: `klz_gatekeeper_session=${gatekeeperPassword}` },
+          validateStatus: null, // Don't throw on error status
+        });
+
+        if (res.status >= 400) {
+          console.error(`❌ ERROR ${res.status}: ${res.statusText} -> ${u}`);
+          hasErrors = true;
+        } else {
+          console.log(`✅ OK ${res.status} -> ${u}`);
+        }
+      } catch (err: any) {
+        console.error(`❌ NETWORK ERROR: ${err.message} -> ${u}`);
+        hasErrors = true;
+      }
+    }
+
+    if (hasErrors) {
+      console.error(`\n❌ HTTP Sitemap Validation Failed. One or more pages returned an error.`);
+      process.exit(1);
+    } else {
+      console.log(`\n✨ Success: All ${urls.length} pages are healthy! (HTTP 200)`);
+      process.exit(0);
+    }
+  } catch (error: any) {
+    console.error(`\n❌ Critical Error during Sitemap Fetch:`, error.message);
+    process.exit(1);
+  }
+}
+
+main();