Content Attribution Checker

The checker reads one public URL and looks for basic signals: page title, meta description, canonical link, robots.txt, llms.txt, copyright language, attribution language, citation policy language, and AI usage policy language. It does not run a browser, log into sites, crawl every page, or search the web for copies.

That narrow scope is intentional. A free first-pass checker should be fast, transparent, and low risk. It can tell you whether your public pages communicate clear expectations. It cannot tell you whether every AI product follows those expectations or whether a particular summary used your work.

A missing signal is not automatically a failure. Many good websites do not have llms.txt yet because the convention is still new. Many blogs rely on visible copyright text without a formal attribution policy. The report is a prioritization tool: it points to easy improvements that make your preferences clearer.

Focus first on gaps that are easy to fix and valuable across many pages. A canonical link and visible copyright line can usually be added in a CMS theme. An attribution policy can be added as a short paragraph. robots.txt and llms.txt can often be published as static files.

Content length is a rough page signal. If a page returns very little text, the checker may have received a blocking page, redirect page, script-heavy shell, or non-article template. That can explain why policy keywords were not found. It can also reveal pages that need better server-rendered metadata for crawlers.

The tool does not fetch images, CSS, JavaScript, or linked pages. It checks the raw HTML response. That makes it lightweight and deployable on Cloudflare Pages Functions, but it also means a fully client-rendered site may show fewer detected signals than a server-rendered or static site.

Use the generated llms.txt and AI attribution policy as drafts. Review them, adapt them to your site, and publish them where they make sense. Then rerun the checker to confirm the public files and visible signals are discoverable. Keep the report as a Markdown note for your publishing checklist.

If you run multiple websites, start with the site that has the most original content or the clearest monetization risk. You do not need a large monitoring platform on day one. A single URL check, a clear policy, and a monthly review habit are enough for the first version.

A lightweight checker reads public HTML and root policy files. That makes it fast and inexpensive, but it also means the result is a signal audit rather than a complete compliance review. It will not know whether a crawler ignored your policy, whether a third-party site copied your article, or whether a JavaScript-only interface hides important metadata from raw HTML.

Use the result as a practical starting point. If the checker finds no title, no canonical link, and no visible attribution language, those are straightforward improvements. If it finds the main signals, your next step is editorial review: make sure the wording reflects your actual licensing and publishing goals. The best report is one that leads to a few clear changes, not a pile of vague anxiety.

The Markdown report is useful because it creates a lightweight record of what the page looked like at the time of review. Save it near your content operations notes, issue tracker, or launch checklist. When you update robots.txt, add llms.txt, or change the attribution page, run the checker again and compare the new report to the old one.

This is also a simple way to prepare for monthly monitoring later. The first version does not store scans, but the report format already matches the future workflow: one URL, a timestamp, detected signals, recommendations, and generated policy text. That structure makes it easy to turn a one-time audit into a recurring review without changing the product promise.