SSL Certificate & Security Headers Checker

Check the SSL/TLS certificate and the HTTP security headers of many domains in one run: one row per domain with the certificate's issuer, expiry date and days left, chain and hostname validity, protocol and cipher, key and signature, SHA-256 fingerprint, the HSTS, CSP, X-Frame-Options and other security headers of the homepage, whether http:// redirects to https://, and an A–F grade with a list of findings. No browser, no proxy, no key.

Price: $1.00 per 1,000 domains ($0.001 per domain checked), plus Apify's default Actor-start event. Domains that fail completely, and blocked ones, are free.

• Certificate monitoring in bulk: daysRemaining, expired, chainValid, hostnameMatch, selfSigned and the issuer for up to 1,000 domains per run. Schedule it weekly and filter on daysRemaining < 30.
• Reports problems instead of failing: expired, self-signed, untrusted and wrong-host certificates are read and reported (the check does not reject them), with the exact OpenSSL reason.
• Security headers with a transparent grade: HSTS (max-age, includeSubDomains, preload), CSP, X-Frame-Options, X-Content-Type-Options, Referrer-Policy, Permissions-Policy, COOP/COEP/CORP and Server/X-Powered-By disclosure. The rubric is printed below; every point lost is a finding with a code.
• Optional TLS 1.0/1.1 test: two extra handshakes per domain show whether a server still accepts the deprecated versions (same price).
• Quick start: press Start. The prefilled input checks crawlee.dev and apify.com (two sites run by Apify): about 3 seconds and $0.002.

Light and polite. Per domain: one TLS handshake, robots.txt, one GET of the homepage (only its headers are read) and one GET of http://host/ to see the redirect. robots.txt is obeyed for the GETs, an honest User-Agent is sent, at most one request per second per host, and blocks (401, 403, 429, 503, challenge pages) are reported as free rows, never worked around.

What it does

For each domain:

1. Guard. Social-media, video and Google hosts are refused before any DNS lookup (platform). The name is resolved and every address must be public: private, loopback, link-local, CGNAT and reserved ranges are refused (ssrf), and the TLS connection goes to that checked address (the host name is sent as SNI), so a name cannot be checked as public and then connect to an internal address.
2. TLS. One handshake to host:port (443 unless set) with SNI and rejectUnauthorized: false: certificate subject, SANs, issuer, validity dates, days remaining, expired and self-signed flags, chain validity against the Mozilla root store (chainValid, chainError), hostname match (tls.checkServerIdentity), negotiated protocol and cipher, ALPN, key type and size, signature algorithm, serial number and SHA-256 fingerprint. Nothing is sent over this connection; it closes after the handshake. A server that only speaks TLS 1.0/1.1 gets one more handshake that allows them, so it is reported as outdated, not as "no TLS".
3. TLS 1.0/1.1 test (optional). Two handshakes pinned to TLS 1.0 and TLS 1.1.
4. Headers. robots.txt of https://host:port, then one GET of https://host:port/. Redirects are followed (at most 5), each hop re-checked by the guard, the platform list and that host's robots.txt. The headers of the final response are parsed; the page body is never read.
5. HTTP to HTTPS. One GET of http://host/ (port 80). Its Location is read: an https:// target means httpToHttps: true (that target is not requested again). Plain-http hops are followed up to 5 times. Port 80 closed means httpCheck: "port-closed" (not a finding).
6. Grade and findings by the rubric below.

Up to 5 domains are checked at the same time (different hosts); requests to one host are spaced at least 1 second apart, longer when robots.txt asks for a Crawl-delay.

Input

Field	What it does
Domains (domains)	Host names or URLs, one per line, up to 1,000 per run. example.org, www.example.org:8443 and https://example.org/any/path all work; the path is ignored and a port in the entry wins over Port. Duplicates (same host and port) are checked once.
Port (port)	TLS port for entries without their own port. Default 443. Headers are read from https://host:port/; the HTTP redirect check always uses port 80.
Test TLS 1.0 and 1.1 (checkLegacyTls)	Default off. On: two extra handshakes per domain fill tls10Accepted and tls11Accepted. Same price.

Example inputs

1. Check your own sites (certificate expiry and headers):

{ "domains": ["crawlee.dev", "apify.com"] }

2. Full audit including deprecated TLS versions:

{
  "domains": ["www.example.org", "shop.example.com", "https://blog.example.net/"],
  "checkLegacyTls": true
}

3. Services on a non-standard port:

{
  "domains": ["mail.example.org", "vpn.example.org:4443"],
  "port": 8443
}

(mail.example.org is checked on 8443, vpn.example.org on 4443.)

Output

One row per domain (shortened; every row has all columns, null when not applicable):

{
  "input": "crawlee.dev",
  "domain": "crawlee.dev",
  "port": 443,
  "status": "ok",
  "billable": true,
  "grade": "B",
  "score": 80,
  "issues": "csp-missing, xcto-missing, referrer-policy-missing",
  "findings": [
    { "severity": "medium", "code": "csp-missing", "message": "No Content-Security-Policy header." },
    { "severity": "low", "code": "xcto-missing", "message": "X-Content-Type-Options: nosniff is missing." },
    { "severity": "low", "code": "referrer-policy-missing", "message": "No Referrer-Policy header." },
    { "severity": "info", "code": "hsts-no-subdomains", "message": "HSTS does not include subdomains." }
  ],
  "ip": "99.86.18.19",
  "protocol": "TLSv1.3",
  "cipher": "TLS_AES_128_GCM_SHA256",
  "alpn": "h2",
  "certSubjectCN": "*.crawlee.dev",
  "certSans": ["*.crawlee.dev", "crawlee.dev"],
  "certIssuer": "CN=Amazon RSA 2048 M04, O=Amazon, C=US",
  "certIssuerOrg": "Amazon",
  "validFrom": "2026-10-04T00:00:00.000Z",
  "validTo": "2027-04-19T23:59:59.000Z",
  "daysRemaining": 175,
  "expired": false,
  "selfSigned": false,
  "chainValid": true,
  "chainError": null,
  "chainSubjects": ["*.crawlee.dev", "Amazon RSA 2048 M04", "Amazon Root CA 1", "Starfield Services Root Certificate Authority - G2"],
  "hostnameMatch": true,
  "keyType": "rsa",
  "keySize": 2048,
  "signatureAlgorithm": "sha256WithRSAEncryption",
  "fingerprintSha256": "E8:B6:1A:EE:00:C5:55:F4:...:1C:8C",
  "tls10Accepted": null,
  "tls11Accepted": null,
  "finalUrl": "https://crawlee.dev/",
  "httpStatus": 200,
  "hstsPresent": true,
  "hstsMaxAge": 31556952,
  "hstsIncludeSubDomains": false,
  "cspPresent": false,
  "xFrameOptions": "SAMEORIGIN",
  "server": "nginx",
  "httpToHttps": true,
  "httpCheck": "redirects-to-https",
  "httpRedirectLocation": "https://crawlee.dev/",
  "errorCode": null,
  "message": null,
  "checkedAt": "2026-10-26T13:40:12.000Z"
}

status is ok (TLS and headers read), partial (TLS read, headers not: errorCode says why) or error (no TLS result). billable says whether the row was charged. The dataset has four views: Overview, Certificate, Security headers and Problems. securityHeaders holds the raw values of the checked headers (each capped at 1,024 characters).

Grading rubric

Start at 100 and subtract per finding: high 20, medium 10, low 5, info 0. Any critical finding makes the grade F. A ≥ 90, B ≥ 80, C ≥ 70, D ≥ 60, F below 60. Without any TLS connection the grade is F; when TLS worked but the headers could not be read (blocked, robots.txt, network), grade is null and only the TLS findings are listed.

Severity	Finding codes
critical	tls-unavailable (no TLS handshake), cert-expired, cert-not-yet-valid, hostname-mismatch, chain-untrusted (self-signed or unverifiable chain)
high (−20)	cert-expires-soon (≤ 14 days), protocol-outdated (negotiated TLS 1.0/1.1), legacy-tls-accepted (TLS 1.0 or 1.1 accepted, when tested), weak-key (RSA < 2048, EC < 256 bits), weak-signature (SHA-1, MD5), hsts-missing, https-downgrade (HTTPS redirects to an http:// page)
medium (−10)	hsts-invalid (no valid max-age), csp-missing, clickjacking-unprotected (no X-Frame-Options DENY/SAMEORIGIN and no CSP frame-ancestors), http-not-redirected (http:// does not redirect to https://)
low (−5)	cert-expires-30d (15–30 days), no-tls13, hsts-short (max-age < 180 days), csp-report-only, xcto-missing, referrer-policy-missing, server-version-disclosed (Server header with a version), x-powered-by-disclosed
info (0)	permissions-policy-missing, cross-origin-headers-missing (COOP/COEP/CORP: only needed for cross-origin isolation), hsts-no-subdomains

The grade is a quick triage aid, not a certification or a full vulnerability scan; other tools weigh things differently.

Error codes

errorCode	Row	Meaning	Charged
invalid-input	error	Not a host name or http(s) URL.	no
platform	error (or partial, after a redirect)	A social-media, video or Google host. Never contacted.	no (yes if TLS of your domain was read)
ssrf	error (or partial)	A private, loopback, link-local or reserved address, or a local name.	no (same)
dns	error	The name does not resolve.	no
connection-refused, timeout, tls-handshake, network, no-certificate	error	No TLS handshake (port closed, no TLS on that port, no answer in 10 s). Graded F.	no
refused	partial	401, 403, 429, 503, 999 or a bot challenge on the homepage. Not retried, not worked around; no further request to that host; after 2 refusals a host is skipped (host-skipped). TLS is still reported.	no
robots-unavailable	partial	robots.txt answered 5xx or 429, or could not be read, twice, or did not answer within the domain's 36 s: the homepage is not requested.	no
robots-blocked	partial	robots.txt disallows / for this Actor: the homepage is not requested.	yes (TLS result)
slow-host	partial	robots.txt asks for a Crawl-delay over 30 s (honoured in full, never shortened, so the GET is skipped).	yes (TLS result)
redirects, headers-too-large, http, timeout, network, dns	partial	More than 5 redirects; response headers over 64 KB; another HTTP problem; the homepage did not answer in time (15 s, or the domain's 36 s ran out); the homepage could not be reached.	yes (TLS result)
internal	error	An unexpected error in this Actor, not a problem with your site.	no

RUN_SUMMARY (key-value store record)

Charged and free rows, grades, TLS handshakes and HTTP requests made, robots.txt bytes read, seconds spent waiting for Crawl-delay, errors by code and stoppedBy: finished, max-charge, unpaid-work-cap, time-limit, migrating, aborted or error.

Pricing

Pay per event: $0.001 per domain checked (event domain-checked), plus Apify's standard Actor-start event. No other fee from this Actor.

Domains with a usable result	Cost
2 (prefilled example)	$0.002
100	$0.10
1,000 (maximum per run)	$1.00

Charged: a row with a TLS result (the certificate was read, whatever its problems) that was not blocked. Free: rows without a TLS result, invalid, platform and private inputs, and rows where the site refused the header check (block or challenge) or its robots.txt could not be read. The TLS 1.0/1.1 test costs nothing extra. Set Max total charge in the run options to cap spending: the run checks the budget before starting each domain and never writes a row it cannot charge.

To keep free work bounded, a run stops (stoppedBy: "unpaid-work-cap") once 100 or more domains gave no usable result and they outnumber the charged ones. Everything found up to then is saved.

Use it via API, n8n or Make

Get an API token in Apify Console under Settings, API & Integrations, and keep it in an environment variable (APIFY_TOKEN below). Never paste it into shared code.

cURL (starts a run, waits and returns the rows; the synchronous endpoint waits up to 5 minutes, so use it for lists of up to a few hundred domains):

curl -X POST "https://api.apify.com/v2/acts/ataism~ssl-security-headers-checker/run-sync-get-dataset-items?token=$APIFY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"domains": ["crawlee.dev", "apify.com"]}'

Python (pip install "apify-client>=3"):

import os
from apify_client import ApifyClient

client = ApifyClient(os.environ["APIFY_TOKEN"])
run = client.actor("ataism/ssl-security-headers-checker").call(run_input={
    "domains": ["crawlee.dev", "apify.com"],
})
for row in client.dataset(run.default_dataset_id).iterate_items():
    print(row["domain"], row["grade"], row["daysRemaining"], row["issues"])

(With apify-client versions before 3.0, call() returns a dict: use run["defaultDatasetId"].)

JavaScript (npm install apify-client):

import { ApifyClient } from 'apify-client';

const client = new ApifyClient({ token: process.env.APIFY_TOKEN });
const run = await client.actor('ataism/ssl-security-headers-checker').call({
    domains: ['crawlee.dev', 'apify.com'],
});
const { items } = await client.dataset(run.defaultDatasetId).listItems();
for (const row of items.filter((r) => r.daysRemaining !== null && r.daysRemaining < 30)) {
    console.log(`${row.domain}: certificate expires in ${row.daysRemaining} days`);
}

n8n:

1. Add a Schedule trigger (for example every Monday).
2. Add the Apify node, operation Run an Actor and get dataset, Actor ataism/ssl-security-headers-checker, input {"domains": ["your-site.example", "shop.your-site.example"]}.
3. Add an IF node on daysRemaining < 30 or grade = F, and send those rows to Slack or email.

Make:

1. Add the Apify module Run an Actor, choose ataism/ssl-security-headers-checker, put your domains in the input JSON and set Run synchronously to Yes.
2. Add the Apify module Get Dataset Items and map the run's default dataset ID (one bundle per domain).
3. Filter on daysRemaining or grade and send the rows to Google Sheets, a ticket system or a webhook.

How it relates to our other Actors

• Web Stack Patch Checker also reads response headers, but only to find the PHP version (X-Powered-By, Server, cookie names) and say whether it indicates a CVE fix. It does not look at certificates or security headers. This Actor reads the security headers and the certificate, and only flags Server/X-Powered-By as information disclosure; it does not interpret versions.
• SEO Page Audit reads page content (titles, meta, headings, schema). This Actor never reads the page body.
• DNS Records Lookup covers the DNS side (A, MX, TXT, SPF, DMARC, CAA). Together with this Actor you get the domain's DNS and TLS posture.

Privacy and terms

• Per domain the Actor makes one TLS handshake (three with the TLS 1.0/1.1 test), reads robots.txt, sends one GET to the homepage and one to http://host/. It reads response headers only, never page bodies, and stores no cookies. Certificates are public by design; they can contain an organisation's name and, rarely, a contact address in the subject, which is passed through as given.
• Nothing is sent to any AI or third-party service. No logins, no proxies, no CAPTCHA solving, no exploit or scan traffic.
• Only check domains you own or manage, or have permission to check. You are responsible for that; the Actor obeys robots.txt for its GETs but does not know your agreement with the site.
• User-Agent: ssl-security-headers-checker-actor/0.1 (https://apify.com/ataism/ssl-security-headers-checker). In robots.txt, User-agent: ssl-security-headers-checker-actor targets this Actor (exact match; otherwise the * group applies).

Limits and honest caveats

• 1,000 domains and 60 minutes per run; at most 36 s of work per domain (waits for the 1-request-per-second limit and Crawl-delay not counted); within that, 10 s for DNS and per TLS handshake, 8 s per TLS 1.0/1.1 probe, 15 s per HTTP answer, 20 s to read robots.txt. A domain that runs out of time keeps its TLS result (errorCode: "timeout", charged) and its http:// check is skipped.
• Only what one connection shows: the certificate served to our SNI from one IP address (the first IPv4 address of the name). Load-balanced or geo-distributed servers can serve different certificates. OCSP/revocation, certificate transparency and the HSTS preload list are not checked (hstsPreload is the directive only). security.txt is not checked.
• The chain is verified against the root store of Node.js (Mozilla's list). A certificate trusted only by an internal or OS-specific root shows chainValid: false.
• Headers come from the final response of the homepage after redirects; other pages of a site can send different headers. A CDN or proxy can add or remove headers.
• The plain-HTTP check uses port 80 and the robots.txt of the host over HTTPS (see STORE.md for why).
• Default memory is 256 MB, plenty for 1,000 domains (measured peak about 110 MB).

FAQ

Is a site with an expired certificate an error row? No. The certificate is read and reported (expired: true, grade F) and the row is charged: that is exactly the result you are looking for.

Why is my site refused? It answered 401, 403, 429, 503 or a bot challenge to our homepage GET. The Actor does not retry or work around that. The TLS columns are still filled and the row is free. Allow the User-Agent above, or check another hostname.

Does it scan for vulnerabilities? No. It reads what any browser sees on the first connection: the certificate and the response headers. The grade is a triage aid.

Does the default run work without setup? Yes. It checks crawlee.dev and apify.com (about 3 seconds, $0.002).

Open-source components

The Actor uses the Apify SDK (apify 3.7.2, Apache-2.0) and Node.js's built-in tls and crypto. The header parser, grading, robots.txt parser and fetching are plain code in this Actor, with no other dependencies.