Update signed manifest to 7.28.5

manifest.json

@@ -1,11 +1,11 @@
 {
-    "version": "7.23.55",
+    "version": "7.28.5",
-    "released_at": "2026-08-13T19:15:44+00:00",
+    "released_at": "2026-09-02T16:09:39+00:00",
-    "package_url": "https://git-cloud.weboria.eu/Weboria/argus-wp-defence/releases/download/v7.23.55/argus-wordpress-defence-7.23.55.zip",
+    "package_url": "https://git-cloud.weboria.eu/Weboria/argus-wp-defence/releases/download/v7.28.5/argus-wordpress-defence-7.28.5.zip",
-    "sha256": "bcab002c465597a3ba70e3db54be6d7ca916979114de51464138b8aea69f8971",
+    "sha256": "f5574faa92ef31247d15f462d67bf7743b3e22b7c3e24fac4e0ccb80708ded38",
     "min_php": "7.4",
     "min_wp": "6.0",
-    "critical": false,
+    "critical": true,
-    "summary": "* Added optional HTTP security headers (helps protect against clickjacking and content-sniffing\n  attacks).\n* Added an optional custom login URL, so your login page isn't sitting at the address every\n  automated attacker already knows to try first.\n* File-integrity monitoring now also watches your .htaccess file, not just wp-config.php.\n* Added optional checkout protection for WooCommerce stores: recognizes scripted/bot checkout\n  attempts and repeated failed orders, a common sign of payment-fraud testing.\n* Improved login protection to recognize credential-stuffing attacks (many different usernames\n  tried from one source), not just simple repeated password guessing.\n* ARGUS Cloud can now recognize when the same attack pattern is seen across multiple protected\n  sites, strengthening protection for everyone connected.\n* Simplified how this site connects to your ARGUS Account -- one-click connect, no code to copy.",
+    "summary": "* Fixed a serious bug where the firewall could block a real visitor (including search engines\n  like Google) on sites using certain page-builder widgets, showing them a \"Forbidden\" page\n  instead of your content -- and could go on to ban their IP address, keeping them locked out.\n* The firewall no longer bans a visitor's IP address after a single suspicious-looking request --\n  it still blocks that one request immediately as before, but only bans the IP once a real\n  pattern of repeated or combined attack behavior is seen. Genuine attacks are still detected and\n  blocked exactly as before; this only prevents an isolated false alarm from locking someone out.",
-    "signature": "rCvDiAjo/+xTQtyrnlK5raUq56pjrx2ylXwJqWPjNRXAkdz3rBLZ9SoQ7ZSrcJ/U0oaw09om/dkiREdXz3g3Aw=="
+    "signature": "tkLviBtuKf9NjiCMOAyLyc3mUH+hvI7zy9CV4B4qUdmkz0m8CqPzYqqIrUxEtL4GFgTtmLMZA1T3w/I1O1CLCg=="
 }