CiteProof – AI Search Visibility, SEO & GEO

No. The SEO modules (meta, schema, sitemap, redirects, breadcrumb, Open Graph, hreflang, IndexNow, local SEO, internal links, AEO score, the /ai-info page, AI crawler control, alt fill at output) work without an account, with no limits.

A free account (created from the plugin in a minute) connects the site to the platform: the page-by-page diagnosis, the Search Console panel, questions tracked in ChatGPT answers, and every month 10 AI drafts, 30 image descriptions and 5 AI rewrites of titles and descriptions. A paid plan raises those to 150, 300 and 200 a month, audits the site on five AI engines (ChatGPT, Perplexity, Gemini, Claude, Google AI) and prepares fixes after every audit that you approve in the plugin. The plugin always tells you how much of the month's quota is left.

Yes. In "automatic" mode the plugin does not print a second title, description, schema, sitemap, Open Graph or IndexNow when one of them is active. You can force any module on or off in CiteProof → SEO. The importer copies your existing meta, templates, redirects and locations if you decide to switch.

A 0–100 index of how easy a page is for an AI engine to extract and cite: sixteen weighted checks (short answer at the top, title with the customer's question, description, questions with answers, subheadings, short paragraphs, lists or tables, enough text, concrete numbers, author with a job title, recently updated, links to other pages and to a source, images described, structured data on, business name in the text, visible to the engines). A check that partly passes earns half; a check that does not apply is left out. It is not a traffic forecast and it is not the dashboard's visibility score.

Only when you press a button. It adds blocks around the content (short answer on top, questions and author box at the bottom, both optional) and fills empty alt attributes at output. "Insert link" wraps one phrase that is already in the text with a link, and "Remove" takes it off; "Use the draft" writes the short answer and questions you approved into their fields. Fixes from the platform are written to the head, robots.txt or llms.txt, never inside a post, and each one can be rolled back.

The page's AEO score and at most three things to do, in plain words, each with the button that does it. Everything technical sits under "Details" (the sixteen checks with their outcome) and "More options" (canonical, robots, internal links, schema, social, languages, locations). The block editor sidebar shows the same list.

With an API key set, the plugin sends to the platform (citeproof.co): the site URL and plugin version, visits from AI crawlers (user agent, path, time), the result of each applied fix, and a content-health snapshot (counts, page ids, titles and URLs, never the text of a page). If you turn on automatic alt text, the URLs of images without alt are sent so a vision model can describe them. IndexNow sends changed URLs to api.indexnow.org. Nothing else is sent and nothing is sent without a key.

No. The Search Console panel reads a summary from the platform, where you connected Search Console with your own Google account. The plugin never asks for Google credentials.

Honestly, not yet. No AI engine has announced it reads the file. The plugin serves it because customers ask for it and it costs nothing; it is not a ranking factor and the interface says so.

Google does not use the IndexNow protocol. Bing, Yandex, Seznam, Naver and Yep do. It still matters for AI visibility because Bing's index feeds ChatGPT search and Copilot.

Two layers. At output, an image without alt gets the media-library alt, the caption, the title or a cleaned file name (a camera name like IMG_4521 yields nothing). If you opt in, images without any alt are sent to the platform in batches of ten and a vision model describes what is visible in your language; the text lands in the media library where you can correct it. There is a monthly quota per account and the count is shown.

Yes. CiteProof → Locations: each one becomes a LocalBusiness node with address, phone, email, opening hours, coordinates and map link, linked to the Organization. Assign a page to a location from the SEO box; show the details with [citami_location], [citami_locations] or [citami_opening_hours].

Fixes generated by the platform are queued in CiteProof → Pending fixes instead of being written to the site. You approve or reject each one. It is on by default; switch it off when you want fixes applied automatically. Every applied fix can be revoked later.

The head output, sitemap, redirects and IndexNow are independent of the builder. The TL;DR, FAQ and author box are added through the standard content filter, so builders that bypass it (some Elementor templates) will not show them: use the shortcodes [citami_tldr] and [citami_faq] inside the builder instead. Caching plugins are fine; the sitemap and llms.txt are served before the cache layer in most setups. The plugin has been tested on its own test harness; please report anything odd on a real site.

The front end adds one small stylesheet on singular pages and a few database reads that WordPress caches. The internal-link index is stored in one option and refreshed on save; the editor panel caches its work per page. Access-log reading and alt generation run in the background.

Deactivate and delete from Plugins. The uninstall routine removes every option, transient, cron event and per-page cache the plugin created. Per-post fields (SEO title, description, TL;DR, FAQ, alt source) stay in the post meta so nothing is lost if you reinstall.

Options are per site and the uninstall routine iterates the network. Each site needs its own API key.