The CITED Score, a Docs Centre, ten more free tools, and a much faster site

Added

• The CITED Score. A single rollup score for how citable your site is to AI answer engines, built from 21 measured signals — brand disambiguation, description accuracy, topical authority, brand radar monitoring, markdown-friendly structure, fake-heading detection and more. It has its own tab on every site, plus a public AI SEO checklist page explaining each check.
• Brand mention and PR tracking. Find unlinked mentions of your brand and press/media coverage across the web, with an alert when new coverage lands. Bundled into your plan, with a per-site switch.
• Docs Centre. Full customer documentation covering every feature, what each paid action costs, and how the automation actually behaves.
• Server log analysis. A real UI for uploaded server logs — see which AI and search crawlers actually hit your site — with retention limits and self-service delete.
• Ten more free tools. AI crawler IP verifier (real reverse-DNS checks), Website Security Score, llms.txt generator, JSON-LD schema generator, any-domain traffic estimate and five more micro-tools, now grouped by category.
• White-label PDF exports. Crawl reports and a one-page GEO battlecard per site, exportable as branded PDFs.
• MCP server expanded from 2 to 10 tools. Drive KinetixSEO from Claude, Cursor or any MCP client.
• Featured Snippet and People Also Ask tracking. Rank tracking now records snippet ownership and PAA presence per keyword, plus the estimated traffic value of each ranking.
• Pause and resume monitoring. Park a site without losing its history.
• Template-page detection. Crawls group pages that share a template, so one fix can be applied to a whole set, and now capture internal anchor text.
• Mobile tap-target checks. Scans flag buttons and links that are too small or too close together on mobile.
• Malicious outbound link detection. Outbound links are checked against Google Web Risk.
• New structured-data rules. ImageObject, Service, QAPage, DiscussionForumPosting, Book and VacationRental, plus validation of image and video sitemap entries.
• A real /pricing page. Full tier comparison at its own URL instead of a redirect to a homepage anchor.

Improved

• Much faster pages. Self-hosted webfonts, overview charts drawn in SVG instead of loading a charting library, guest pages served a stylesheet scoped to what they can actually reach, markdown rendered server-side, pre-compressed build output and a smaller OG image. Fewer bytes, less JavaScript, faster first paint everywhere.
• AI crawler coverage 28 → 54 bots. Plus ai.txt and TDMRep support, and an alert when your CDN starts blocking AI crawlers.
• Honest scoring. The CITED Score no longer reports unmeasured checks as passes, E-E-A-T no longer awards authority without independent reputation evidence, and scan scores stay comparable across scoring-version changes.
• SERP truncation measured in pixels. Title and description truncation is predicted the way Google actually renders it, and canonical conflicts are predicted before they bite.
• Clearer costs before you click. Every charged action has a documented cost basis, prices are shown in real euros, and the run-now button warns when your balance is short.
• Rank-tracking honesty. "Domain does not rank" is recorded as a real result instead of a provider failure, and unranked days are no longer hidden.
• Article publishing. Published articles go out on a jittered mid-morning schedule, with redesigned OG images and post-publish lint rules.

Fixed

• Scan false positives. Long pages are no longer accused of scaled content abuse, ccTLD neighbours are no longer treated as your own domain, interstitial detection is advisory and now catches JS-injected overlays, and schema advice is calibrated to 2026 rich-result reality.
• Sites workspace. The Add keyword and Add competitor forms, keyword delete, and the Search Console unlock banner all work again; the visibility chart anchors to the latest data point instead of your browser clock.
• Unreachable sites. Scanning a site that is down now explains what happened instead of failing opaquely.
• PageSpeed reliability. Transient Lighthouse failures are retried rather than failing the whole scan, with timeouts that match real run durations.
• Keyword gap accuracy. Per-competitor provider failures are isolated, real search volume is pulled for backlog ranking, and comparison-intent keywords are called out separately.
• Article quality gates. Question words no longer skew the cannibalization check, fabricated internal links are caught before publish, and fix packets always carry their entity list.
• Marketing copy accuracy. The GEO score described on the marketing pages is the score the product actually computes, before/after examples use real production scans, and two features that could not run are no longer offered.