Paste what you would put in a bot profile. You get back what is already there, and the lines to add.
Nothing you paste leaves this browser.
No. No requests except this page's own files, no analytics, no cookies, no service worker. Two local storage keys remember your language and your theme, nothing else. It works offline once loaded.
{n} English charters from 86 public GitHub repositories, snapshot 2026-09-04, every source pinned to a commit sha. It is not a census of public templates.
2 of the 8 rules cleared the publication gate against blind hand labels. The rest are reported only from a sample of 60 on the method page. The secret-shape rule is left out on purpose: its count at this corpus carries no information.
No. There is no grade and no total. It leads with what the charter already does, then hands you lines for what is missing.
The interface is translated, the eight detectors read English only. Text that does not read as English gets a notice and the results are shown anyway.
No. This reads text, not behaviour. A charter with an approval line can still be run without one, and one without it can be run carefully.
It still runs. The word count says so, because the audit corpus starts at 40 words and the comparison sits on thinner ground below that.
Yes. The rules are regular expressions over normalized English text. An approval gate phrased in a way the corpus never used will be missed, which is why nothing here is scored.
No. One person built this. It reads text people published themselves and never touches x.ai.
This reads text, not behaviour. The corpus is {n} English charters published in public repositories on 2026-09-04. 6 rules are measured but not shown here, and the method page says which and why.