Privacy

Everything the plugin records stays in one table in your own WordPress database. Rows are deleted after 90 days; you can choose anything from 7 to 730. Nothing about your site, your visitors or the report is sent to us or to anyone else.

Visitors referred by AI engines: the plugin stores the time, the engine and the page they landed on. It stores no IP address, no user agent and no cookie, and sets nothing in the browser.

AI bot hits: the plugin stores the bot's IP address and user agent so the hit can be checked against the bot's published IP list.

Outbound requests: once a day, through WP-Cron, it downloads the public IP lists published by OpenAI, Anthropic, Perplexity, Amazon, Common Crawl, Mistral and DuckDuckGo. These are plain GET requests that carry nothing about your site beyond what any web request does. For Amazonbot and CCBot it can also run reverse and forward DNS lookups through your server's normal resolver. Those are the only outbound requests it makes.

Uninstalling the plugin deletes its table and settings.

A private beta, by invitation: a script tag and a crawl reporter (a Cloudflare Worker, or a PHP or Node snippet) that send a site's AI traffic here instead of to a WordPress database. The plugin doesn't use it and sends nothing.

Visitors referred by AI engines: the script tag sends one beacon, only when the visitor came from an AI engine, with the page's path and the engine. We keep those two things and when the beacon arrived. The tag sets no cookies and uses no storage. To limit made-up beacons, the collector counts them per visitor for a day under a one-way hash of the IP address with a salt that changes daily, in a temporary cache only; the IP address itself is never stored.

AI bot requests: the crawl reporter sends a report only for requests whose user agent names an AI bot: the path (never the query string), method, status, cache status, a request ID, the time, the user agent and the IP address. The IP is checked on arrival against the bot's published IP list, or its reverse and forward DNS. If it's confirmed, we keep the IP, user agent and the evidence (the list's version, or the DNS answers). If it isn't, we keep the bot's name, the request's details and "claimed", with no IP or user agent.

Audit mode: for checking the tracker itself, the site's owner can have audit mode turned on for one site. While it's on, the IP addresses of unconfirmed requests are also kept, encrypted, in a separate table, so they can be checked independently against the bots' lists. That table's rows are deleted automatically 7 days after audit mode is turned off. The dashboard shows when it's on.

How long: requests, referrals and the collector's daily counts are deleted after 90 days; the daily jobs' run records after 120.

Logs: requests to the collector and the script (/c/ and /t/) are kept out of this site's application log and its web server's access and error logs. Errors there go to the log and our error reports with the route and the reason only, never the IP, user agent, headers or body, and these requests aren't counted as page views.

Outbound requests: to check bots, the collector downloads the same official IP lists the plugin does, once a day, and runs reverse and forward DNS lookups for Amazonbot and CCBot.

The dashboard: signing in sets the same session and security cookies the waitlist uses. Your account is your email address and a hashed password.

No ads, no third-party trackers. Fonts and scripts are served from this site, not from third parties.

Visit counts: the server counts page views and plugin downloads with our own self-hosted, cookieless analytics. It records the page, the referring site, your country and device type, and a visitor code that is a one-way hash of your IP address and browser, changes every day and can't be traced back to you. Your IP address itself is never stored, and if your browser sends Do Not Track or Global Privacy Control, the visit is counted with no visitor code at all. Joining, confirming or leaving the waitlist is counted the same way, without your email or site address. None of this is shared or sold.

Cookies: one session cookie and one security token cookie, which the waitlist form needs to work. They contain no tracking and expire after two hours.

The waitlist: if you join, we keep your email address, your site's address, how many sites you said you run (if you did), and when you signed up and confirmed. We use them for one thing: to email you when the hosted tracker opens to more sites. We don't sell or share them. To send the confirmation email we use SendWP, which handles the message only to deliver it.

Rate limits: to stop abuse, the waitlist form counts sign-ups per IP address and per email for up to a day. The count is kept in a temporary cache, not with your sign-up.

Leaving: every email we send has a link that deletes your address at once. If you never confirm, we never email you again unless the form is filled in again.

Server logs: the web server keeps standard access logs (address, time, page requested) for operating and securing the site.