PAUL BUNYAN
Diamond Sierra Site Defense
Diamond Sierra Site Defense checks your WordPress site once a day and tells you, in plain English, what is wrong and how to fix it. A daily malware scan of your files and your live homepage, one-click fixes you can undo, a firewall that can run before WordPress loads, and login protection are all in the free plugin, with no account to sign up for and no firewall rules held back for paying users.

Daily malware scan
- Every WordPress core file is compared with the official checksums from WordPress.org.
- Plugins and themes from the WordPress.org directory are compared with their official copies, so a tampered file (a supply-chain attack) is caught and verified files are skipped.
- A signature scan of themes, plugins, must-use plugins and uploads looks for web shells, obfuscated
eval()backdoors, droppers and hidden-admin code. - It catches ClickFix and fake CAPTCHA pages that trick visitors into running commands on their own computer, clipboard hijacking scripts, and loaders that pull code from blockchain smart contracts.
- Your homepage is fetched like a visitor arriving from Google, and again as Googlebot, to catch injections and SEO spam that hide from logged-in users.
- New and changed PHP files are tracked between scans.
- Accounts are checked for what they can do, not what their role is called, so a “customer” with administrator powers is reported, and so are the administrator names known malware makes up.
- Plugins that aren’t from WordPress.org are listed, and reported as critical when they also show signs of malware, such as hiding from the Plugins screen or creating administrators.
- Malware that reinstalls itself is caught: code that turns a plugin back on when you turn it off, watchdog files, the settings and scheduled tasks it keeps in the database, and hostile
.htaccessfiles. - Signatures for the 2025 and 2026 WordPress malware campaigns published by GoDaddy, Patchstack and Sansec, among them DollyWay, the fake WooCommerce security patch, the FunnelKit checkout skimmer and the SocGholish fake browser update.
- Copies of
wp-config.php, database dumps, site backups, error logs,.envfiles and a.gitfolder left in the website folder are reported, together with whether a visitor could download them. PHP errors that show your server path to visitors are reported too. - After every scan the dashboard shows what changed since the one before: the new issues by name, and how many were resolved.
- Risky plugin alerts with no setup and no API key: plugins removed from WordPress.org for a security issue, security updates waiting, and plugins abandoned for over two years.
- The scan runs in small steps, so it works on shared hosting without timeouts.

Fixes you can undo
- Repair puts back the verified copy of a changed WordPress, plugin or theme file from WordPress.org, and View differences shows exactly what was changed first.
- Delete file removes a file that isn’t part of WordPress, a plugin or a theme. A copy is kept for 30 days on the Quarantine tab, in case something needed it. Official files can’t be deleted, only repaired.
- A verification scan runs after every fix to confirm the malware is gone.
- View file shows the code with the matched lines marked, before you decide. The database login and keys in
wp-config.phpstay hidden. - For malware that restores itself: Remove tasks and Delete settings clear what it keeps in the database, and Remove admin rights or Delete account deal with the fake administrators it creates.
- Mark as fixed for a finding you dealt with by hand.
Tasks and hidden code
- Every scheduled task WordPress will run, with the plugin or theme behind it. Tasks that belong to a known self-restoring malware family are flagged, and tasks nothing uses any more can be deleted.
- Must-use plugins and drop-ins, which run on every page without appearing on the Plugins screen, are listed with what the scan found in each.
Firewall
- Blocks SQL injection, cross-site scripting, path traversal, remote file inclusion, command injection and PHP object injection, in forms, cookies and JSON sent to the REST API.
- Blocks executable and disguised PHP uploads, and SVG files with scripts in them.
- Blocks vulnerability scanners and probes for
.env,.git,wp-configbackups and known shell names. Repeat offenders are blocked automatically. - An address hunting through dozens of missing pages for vulnerable plugins and old backups is blocked. Missing images and styles don’t count.
- Rate limiting, with separate limits for browsers and for bots, that slows a visitor down or blocks them. Logged-in users, real search engines and the WooCommerce cart and checkout are never counted.
- Fake search engines are blocked: a bot calling itself Googlebot, Bingbot or Applebot is checked against the address lists Google, Bing and Apple publish, and then against DNS.
- Bad bots are turned away: site copiers, email harvesters, scraping tools and form posts with no browser headers. SEO tool crawlers and AI training crawlers can be turned away too.
- Your own block list of IP addresses and ranges, user agents, referring sites and visitor hostnames. A save can never block the administrator making it.
- Comment spam bots are caught without a CAPTCHA, with an invisible decoy field and a time stamp, and go to the Spam folder.
- An optional REST API lock for logged-out visitors, with the routes front-end features need left open.
- New installs spend their first week in log-only mode, so you can see what would be blocked before anything is.
- Logged-in administrators and editors are never filtered, so page builders keep working.
- Early firewall, one click: the firewall runs before WordPress and your other plugins load, so an attack on a vulnerable plugin is stopped before that plugin’s code even starts. Site Defense sets it up itself in
.user.ini,.htaccessor, where the server allows neither,wp-config.php. It works next to another firewall that already runs this way, such as Wordfence’s: Site Defense runs first, then hands over. - A private unblock link for your password manager. If your own address ever gets blocked, opening it removes the block and trusts that address for an hour. It gets nobody past your password or two-factor login.

Login protection
- Two-factor login with any authenticator app, or with codes sent by email to your account email, the site admin email or another address you choose. One-time backup codes can be copied or downloaded.
- CAPTCHA on the login, registration, lost-password and comment forms: a built-in math question with no account needed, Cloudflare Turnstile, Google reCAPTCHA v2 or hCaptcha. A CAPTCHA outage or a wrong key never locks you out.
- Passwords found in known data breaches are refused, using Have I Been Pwned; the password itself never leaves your site.
- A custom login address, a honeypot that stops login bots, and limits on failed logins per address and per account. Addresses an account has used before are never locked out.
- Strong passwords for administrators, editors and shop managers, checked wherever a password is set. Trap usernames like “admin” lock bots out on the first try and can’t be registered.
- An optional login allowlist: powerful accounts can only log in from the addresses you list, while customers log in as before.

Email alerts
By default you get one security email a week, on Monday morning, and only when something is new. You can switch to a daily email, or one straight after every scan. A new administrator and Site Defense being switched off are always emailed straight away.
- The scan found a serious problem, or an all-clear after every scan if you want one.
- An administrator signed in from a new address. This one is skipped while Wordfence already emails every administrator login.
- A new administrator account was created, or someone was promoted to administrator.
- Diamond Sierra Site Defense itself was deactivated, which malware often does first.
- Plugin, theme and WordPress updates are waiting.
- How many addresses and accounts were locked out after failed logins, if you want to know.
Hardening
A security score explains every check, and each one has its own Apply button: turn off XML-RPC, the file editor, application passwords and username lookups; block PHP in uploads and directory listing; add security headers and hide version numbers, including the PHP version. A file permissions check finds files that others can write to, and fixes them in one click.

Works with the security plugin you already have
Run only one firewall. When another security plugin is active, Diamond Sierra Site Defense pauses its own firewall and login limits, so the same visitor is not blocked twice, and keeps scanning. Results also appear under Tools › Site Health, and WP-CLI commands get you back in after a lockout: wp site-defense unblock, reset-login-url and 2fa-reset.
Get Diamond Sierra Site Defense
Diamond Sierra Site Defense is free on WordPress.org. To install it, go to Plugins › Add New in your dashboard and search for “Diamond Sierra Site Defense”. It needs WordPress 5.8 or later and PHP 7.4 or later, and on a multisite network it is network-activated to protect every site.
How to use each feature, and how to get back in if you are ever locked out, is in the documentation. What changed in each version is in the changelog.
What it sends to other services
- WordPress.org: the names and versions of WordPress, your plugins and your themes, to fetch official checksums and clean copies of files. This is the same information WordPress sends when it checks for updates.
- Google, Bing and Apple: once a day the plugin downloads the lists of addresses their crawlers use, to tell real search engines from fake ones. Like every download WordPress makes, the request includes your site’s address.
- Have I Been Pwned: only if you turn on breached-password blocking. The first 5 characters of the password’s SHA-1 hash are sent; the password and the rest of the hash never leave your site.
- Cloudflare Turnstile, Google reCAPTCHA or hCaptcha: only the one you choose, on the forms it protects.
- Google Safe Browsing: only if you turn on the blocklist check. Your site’s address is sent once per scan.
