Site Compliance Checker

Check a list of public websites and flag likely issues. Each site comes back as one row, with a score, a short list of findings, and a shareable HTML report.

Not legal advice; general information only.

These findings flag likely issues. This is not legal advice or a compliance certification.

Who it is for

Site owners, agencies and consultants who want a quick, repeatable look at cookies set before anyone clicks, accessibility issues, AI crawler readiness, and basic privacy signals.

What you get

• One result per website, even when several pages are opened.
• Cookies set before any click: name, domain, expiry, and whether the cookie is first party or third party. Cookie values are never saved.
• Known third party tracker domains that the page contacted, and whether a consent banner appears to be present. When it does, the result names the signals that matched. A script from a consent tool, on its own, is not reported as a banner. Cookie values are not among those signals.
• An accessibility scan: how many issues were found by severity, the main rule ids, and a few example selectors. Page text is not saved.
• AI readiness: robots.txt rules for well known AI crawlers, whether llms.txt is present, whether a sitemap was found, and meta robots or noai tags.
• Privacy signals: HTTPS, a privacy policy link, and which common security headers are present. Header values are not saved.
• A score from 0 to 100 and a letter grade, plus a short list of likely issues. If a check does not finish, the grade is incomplete and there is no score. That is not a pass.
• An HTML report saved with the run.
• An optional client-ready report you can hand to a client. Turn it on under Client-ready report. Off leaves the run unchanged.

Private and internal addresses are skipped and are not charged, including a redirect or a file the page itself tries to load. A site that could not be checked is saved with no score and is not charged. That includes a site whose robots.txt answered when the browser could not open a page. A robots.txt rule that tells this checker not to open pages is still a completed check. Sites that time out are not charged. The checker does not fill in forms, click buttons, or type into the page.

Tutorial

1. Paste public website URLs into Website URLs. One result is returned per website.
2. Leave Max pages per site at 3 unless you need a wider look. The price is per site, not per page.
3. Set Max sites to the budget you want for this run. The default is 5 and the maximum is 25.
4. Leave Respect robots.txt on unless you operate the site and you want pages opened anyway.
5. Leave the four checks on, or turn off any you do not need. At least one check has to stay on.
6. Run the Actor.
7. Open the dataset for one row per site. Open the OUTPUT record for the summary, and the REPORT record for the HTML report.
8. Schedule the task if you want a standing check. A run that checks nothing still pays only the small start fee.

Client-ready report

Use this when you want one file to send to a client. It is off by default. Off does not add a file and does not add a fee.

1. Paste the websites you want to check.
2. Set Client-ready report to HTML, or to PDF if you also want a PDF of the same page.
3. Run the Actor.
4. Open the Output tab. Client report (HTML) is REPORT.html in the key-value store. Client report (PDF) is REPORT.pdf when you chose PDF. Client report summary lists those keys.

The page has a cover with the site, the date and an overall grade, a short summary in plain language, and a table of findings. Each finding says what it means and how to fix it. A run with more than one site has a section per site. Pass, warning and fail stay readable in black and white. The footer says Generated with Site Compliance Checker.

The report fee is charged once per run, and only after the file is saved. HTML and PDF together are one fee. A failed save is not charged. The dataset and the existing REPORT summary stay included in the site fee.

Example input

{
  "startUrls": ["https://example.com/", "https://crawlee.dev"],
  "maxPagesPerSite": 3,
  "maxSites": 5,
  "concurrency": 1,
  "respectRobotsTxt": true,
  "includeCookies": true,
  "includeAccessibility": true,
  "includeAiReadiness": true,
  "includePrivacySignals": true
}

Example output

One dataset row per website that was actually checked:

{
  "url": "https://example.com/",
  "score": 72,
  "grade": "C",
  "disclaimer": "These findings flag likely issues. This is not legal advice or a compliance certification.",
  "pagesChecked": 3,
  "consentBannerLikely": true,
  "consentVendor": "OneTrust",
  "consentSignals": ["cookie:OptanonConsent", "global:OneTrust", "script:cdn.cookielaw.org"],
  "thirdPartyCookies": 2,
  "trackerDomains": 1,
  "criticalAccessibility": 0,
  "likelyIssues": [
    { "severity": "medium", "message": "Third-party cookies were set before any interaction." }
  ]
}

Cookie rows inside the result list the cookie name, domain, expiry and party. They never include the cookie value.

Pricing

You pay a small fee when the run starts, plus a fee for each public website that is checked.

What	Price
Run start	$0.00005
Site checked	$0.04
Client-ready report	$0.04

The site fee covers every page opened for that site, up to Max pages per site. Private addresses are not charged. A site that could not be checked is not charged, including when robots.txt answered but no page opened. A robots.txt block is still charged, because that check finished. A site whose grade is incomplete, because a check did not finish, is not charged. Sites that time out are not charged. The score, the findings and the HTML summary in REPORT are included. The client-ready report is charged once per run, only when that file is saved. Leave Client-ready report on Off and that fee is not charged. Add $0.04 to any example below when the report is saved.

Examples

• 1 site: about $0.04005
• 5 sites, the default: about $0.20005
• 25 sites, the maximum: about $1.00005
• A scheduled run that finds nothing new to check: $0.00005

Privacy

The Actor does not collect personal data. It does not store cookie values, form contents, names, email addresses, phone numbers, or IP addresses. Page text and security header values are not saved. Logs are scrubbed the same way. Please do not put personal data in the input URLs. Query values that look like addresses, tokens or passwords are removed before a URL is stored.

Limits

• At most 25 websites per run, and at most 10 pages on each website.
• At most 2 websites open at the same time. The default is 1. Choose 2 when the run has at least 4 GB of memory. The sites share one browser. Each site gets its own browser context, and that context is closed when the site finishes or runs out of time.
• The default memory is 2 GB. A heavy page plus the accessibility scan needs more than 1 GB.
• Each site has a time limit of 45 seconds plus 20 seconds per page, and never more than 180 seconds. The limit covers the file fetches and the browser. A site that runs out of time is saved as incomplete and is not charged. One slow site does not hold up the others.
• No new site starts in the last 60 seconds of the run. The summary then says the run stopped early. Sites already saved stay saved.

Next additions

• Agency white-label logo
• Dutch report