Skip to main contentSEO for AI Agents.
Changelog

What we shipped, when.

Product releases, new checks, methodology revisions, and content drops. Updated every time we ship.

2026.5

Competitor comparisons + content depth

  • content
  • seo
  • Shipped /vs landing pages for Seobility, Ahrefs, Semrush, Screaming Frog, and Moz with capability matrices.
  • Expanded the blog from 12 to 24 long-form posts and the glossary from 12 to 40 terms.
  • Added BreadcrumbList and Article JSON-LD to every post and term for richer SERP rendering.
2026.4

AI-search wedge positioning

  • positioning
  • seo
  • Rewrote the homepage around AI Overview, ChatGPT Search, Perplexity, and Claude citation intent.
  • Launched /ai-visibility-checker, /get-cited-by-chatgpt, and /seo-checker as long-form landings.
  • Every numeric claim on marketing pages now carries an inline citation to a primary source.
2026.3

Self-critiqued AI briefs

  • ai
  • audit
  • Audit briefs are now drafted by Gemini 2.5 Pro, then re-scored by a critique pass against a rubric.
  • Headline metric is selected for counter-intuition, not just severity.
  • Action plan now ships with 30 / 60 / 90 sequencing and owner suggestions.
2026.2

GEO + entity check set

  • geo
  • checks
  • Added Organization @id + sameAs validation against Wikidata edges.
  • Live brand-visibility probe checks whether the site appears in top results for its own brand.
  • Author Person entity validation across long-form content.
2026.1

AIO check set + llms.txt parser

  • aio
  • checks
  • llms.txt detection and validation against the proposed spec.
  • Answer-first lead heuristic for top-of-page content.
  • Question-shaped H2/H3 ratio + factual density per 1000 words.
2026.0

Evidence-receipt drawer

  • audit
  • ux
  • Every finding now exposes the raw JSON the check evaluated.
  • Receipts include HTTP headers, JSON-LD blocks, affected URLs, and heading trees.
  • Click any finding to expand the receipts panel inline.
2025.12

Render-vs-HTML diff

  • technical
  • Added a plain (no-JS) fetch of the entry page and diffed it against the JS-rendered markdown.
  • Flags client-side-only content invisible to engines that don't render JS.
2025.11

Monitoring + diffs

  • monitoring
  • Monitored sites re-run on a weekly or daily cadence with score diff and finding-level changelog.
  • Email alerts when a category score drops more than ten points.

Want the methodology behind these checks?

Every check is documented in the methodology, with its weight, severity, and the primary source it derives from.

MethodologyPrimary sources