| 开发者 | rankpresscom |
|---|---|
| 更新时间 | 2026年7月13日 16:41 |
| PHP版本: | 7.4 及以上 |
| WordPress版本: | 7.0 |
| 版权: | GPLv2 or later |
| 版权网址: | 版权信息 |
No. The AI visibility signals are generated locally on your site, with no key and no account. The optional hosted AI help is included by Rank Press and also needs no key.
It is a public page at /brand-facts/ (plus a machine-readable .well-known/brand-facts.json) that presents verified facts about your business: what you do, since when, where, your offerings and your key facts. AI engines can read and quote it to answer questions about you accurately, instead of inventing details. You fill your business profile once and the page stays in sync.
AI engines read structured, machine-friendly signals to decide what a site is about, whether the source is credible, and whether to quote it. This plugin publishes the formats they look for (llms.txt, Schema.org, E-E-A-T entity data, a Brand Facts page and AI robots signals) so AI engines can read, trust and cite your pages and facts.
Yes. You describe your organization and your author once (role, expertise, credentials, areas of knowledge, professional links), and the plugin turns that into Schema.org Person and Organization data and includes it on your Brand Facts page, so engines see a credible, identified source behind your content.
No. The files are generated from your existing content and cached. There is no third-party script on your front end and no external request for the local features.
No. It adds AI-specific signals and serves its sitemap at its own path, so it does not duplicate or override anything you already run.
The local features send nothing. Only when you click the optional "Analyze my site" button, and after you consent, your site URL is sent to the Rank Press hosted AI, which reads your homepage and returns suggested brand details. See the External services section below.
Yes. One toggle refuses known training crawlers in robots.txt while keeping citation and search bots allowed, so you stay visible in AI answers without feeding training datasets.