WebForensik

A Content Security Policy (CSP) is a doorkeeper rule for the browser: "Scripts and styles may only be loaded from these allowed sources." Without CSP, injected malicious code (XSS) can freely fetch anything. Start with a simple, secure baseline.

Apache server (classic hosting at most providers)

File: .htaccess in the web root

<IfModule mod_headers.c>
    Header always set Content-Security-Policy "default-src 'self'; img-src 'self' data: https:; style-src 'self' 'unsafe-inline'; script-src 'self' 'unsafe-inline'; font-src 'self' https: data:; object-src 'none'; frame-ancestors 'self'; base-uri 'self'"
</IfModule>

⚠ This policy is intentionally pragmatic (allows inline styles since many themes/plugins rely on them). If something breaks after enabling: F12 → Console shows "Refused to load…" — add the affected domain after script-src / img-src.

# BEGIN WebForensik CSP
<IfModule mod_headers.c>
    Header always set Content-Security-Policy "default-src 'self'; img-src 'self' data: https:; style-src 'self' 'unsafe-inline'; script-src 'self' 'unsafe-inline'; font-src 'self' https: data:; object-src 'none'; frame-ancestors 'self'; base-uri 'self'"
</IfModule>
# END WebForensik CSP

add_action('send_headers', function () {
    header("Content-Security-Policy: default-src 'self'; img-src 'self' data: https:; style-src 'self' 'unsafe-inline'; script-src 'self' 'unsafe-inline'; font-src 'self' https: data:; object-src 'none'; frame-ancestors 'self'; base-uri 'self'");
});

✓ How to verify it works: Open page, F12 → Console — no red "Refused to load…" messages. Network tab → first request → Response Header "content-security-policy" visible.

Without a Referrer-Policy the browser sends the full URL of your current page (incl. search terms, usernames in URL params) on every click to the destination site. Privacy-relevant per Art. 5 GDPR. Recommended safe setting: strict-origin-when-cross-origin.

Apache server (classic hosting at most providers)

File: .htaccess in the web root

<IfModule mod_headers.c>
    Header always set Referrer-Policy "strict-origin-when-cross-origin"
</IfModule>

⚠ "strict-origin-when-cross-origin" is the modern standard: on cross-domain clicks only your origin (no path/params) is sent — internal clicks include the full URL. Stricter is "no-referrer" (send nothing) but it breaks some analytics tools.

<IfModule mod_headers.c>
    Header always set Referrer-Policy "strict-origin-when-cross-origin"
</IfModule>

add_action('send_headers', function () {
    header('Referrer-Policy: strict-origin-when-cross-origin');
});

✓ How to verify it works: F12 → Network → first request → Response Headers — "referrer-policy: strict-origin-when-cross-origin" must be present.

Without clickjacking protection your site can be invisibly embedded into a malicious page ("enter your password here" — the click actually lands on your overlayed login form). Fix: set SAMEORIGIN (only your own domain may embed).

Apache server (classic hosting at most providers)

File: .htaccess in the web root

<IfModule mod_headers.c>
    Header always set X-Frame-Options "SAMEORIGIN"
    Header always set Content-Security-Policy "frame-ancestors 'self'"
</IfModule>

⚠ Use both headers: X-Frame-Options for older browsers, frame-ancestors for modern ones. If you already have a CSP, add "frame-ancestors 'self'" there — don’t duplicate.

<IfModule mod_headers.c>
    Header always set X-Frame-Options "SAMEORIGIN"
</IfModule>

add_action('send_headers', function () {
    header('X-Frame-Options: SAMEORIGIN');
});

✓ How to verify it works: F12 → Network → Response Header: "x-frame-options: SAMEORIGIN".

Permissions-Policy controls whether scripts (including third-party) may access camera, microphone, location, motion sensors etc. GDPR-relevant because sensitive device APIs can otherwise be reached unnoticed.

Apache server (classic hosting at most providers)

File: .htaccess in the web root

<IfModule mod_headers.c>
    Header always set Permissions-Policy "camera=(), microphone=(), geolocation=(), payment=(), usb=(), accelerometer=(), gyroscope=(), magnetometer=(), interest-cohort=()"
</IfModule>

⚠ "()" at the end means: no caller (not even your own page) may use this API. If you need geolocation (e.g. a map feature): use geolocation=(self) instead of geolocation=(). "interest-cohort=()" disables Google’s FLoC tracking.

<IfModule mod_headers.c>
    Header always set Permissions-Policy "camera=(), microphone=(), geolocation=(), payment=(), usb=(), interest-cohort=()"
</IfModule>

add_action('send_headers', function () {
    header('Permissions-Policy: camera=(), microphone=(), geolocation=(), payment=(), usb=(), interest-cohort=()');
});

✓ How to verify it works: F12 → Network → Response Header: "permissions-policy" visible.

GDPR-relevant: visitor data (at least IP + User-Agent) is transmitted to servers outside the EU/EEA. Since the Schrems-II ruling (2020) this requires Standard Contractual Clauses + supplementary technical measures AND prior consent. Best fix: replace with EU alternatives where possible.

✓ How to verify it works: Load in incognito mode, F12 → Network → list all requests → check "Domain" column for non-EU servers. After migration no unrequested US domains should load.

Only some of your external resources (0 of 2) are protected by SRI. Add integrity attributes to the remaining ones too.

✓ How to verify it works: F12 → Console on page load: no "Failed to find a valid digest" messages. Source: all external <script>/<link> have an integrity attribute.

SRI (Subresource Integrity) is a checksum in HTML that defines what an externally loaded file MUST look like. If someone tampers with the external file (e.g. a CDN gets compromised), the browser refuses to load it. You add the "integrity" attribute on the script/link tag.

✓ How to verify it works: F12 → Network → requests with status 200 from CDN domains (cdn.jsdelivr.net, cdnjs.cloudflare.com etc.) → in HTML source the tag must contain "integrity=\"sha384-…\" crossorigin=\"anonymous\"".

CAA records (Certification Authority Authorization) define in DNS which Certificate Authorities are allowed to issue certificates for your domain. Without a CAA record an attacker could request a fraudulent certificate for your domain at any CA. CAA is pure DNS configuration — set in your registrar/DNS-panel, NOT in WordPress.

Name   Type   Flag   Tag      Value
@      CAA    0      issue    "digicert.com"
@      CAA    0      issue    "letsencrypt.org"
@      CAA    0      iodef    "mailto:security@your-domain.com"

✓ How to verify it works: On https://www.ssllabs.com/ssltest/analyze.html?d=your-domain.com → "DNS CAA" section → all your CAs should be listed. Or via dig: dig CAA your-domain.com.

Your domain has no IPv6 address (AAAA record). Over 40% of users (especially mobile) reach the internet via IPv6 — they must take the slower IPv4 gateway detour.

✓ How to verify it works: dig AAAA your-domain.com — or online https://ipv6-test.com/validate.php?url=your-domain.com.

SPF (Sender Policy Framework) defines in DNS which servers may send emails on behalf of your domain. Without SPF any phisher can spoof emails from you — and recipients are more likely to fall for them.

✓ How to verify it works: dig TXT your-domain.com | grep spf — or online https://www.kitterman.com/spf/validate.html.

DMARC combines SPF and DKIM into an explicit instruction for receiving mail servers: "What to do if emails claim to come from us but SPF/DKIM fail?" Without DMARC each server decides — usually generously. With DMARC=reject you effectively prevent phishing in your name.

✓ How to verify it works: dig TXT _dmarc.your-domain.com — or online https://dmarcian.com/dmarc-inspector/.

A security.txt (RFC 9116) tells security researchers how to responsibly report vulnerabilities to you. Without it, reports may go to spam or never be sent. A plain text file at the correct path is enough.

Apache server (classic hosting at most providers)

File: /.well-known/security.txt (create the folder if it doesn’t exist)

Contact: mailto:security@your-domain.com
Expires: 2027-12-31T23:59:59.000Z
Preferred-Languages: en, de
Canonical: https://your-domain.com/.well-known/security.txt

⚠ Replace "security@your-domain.com" with your actual security contact (or a generic info@). "Expires" must be a future date and should be renewed regularly. The file is plain .txt, not PHP.

Contact: mailto:security@your-domain.com
Expires: 2027-12-31T23:59:59.000Z
Preferred-Languages: en, de
Canonical: https://your-domain.com/.well-known/security.txt

✓ How to verify it works: Open https://your-domain.com/.well-known/security.txt in a browser — content must be visible (no 404).

The privacy policy link returns an error (HTTP HTTP/1.1 404 Not Found). Effectively the same as no privacy policy — same legal status as missing.

✓ How to verify it works: Privacy link in footer → opens the page with status 200, content visible.