Do I need an AuditAE account to use the plugin?
No. The AI Readiness Score, the AI Crawler Tracker and robots.txt audit, the llms.txt generator, Organization schema, the per-post Schema box (Article, Blog post, HowTo, Product) and FAQ schema box, the broken link checker, orphaned content and internal link suggestions, Redirects, image file renames, IndexNow and Query Fan-Out coverage are all free and standalone. They work the moment the plugin is activated, and pairing doesn’t take any of them away. The Connect to AuditAE tab is optional and only relevant if you want the paid citation audits, AEBOT, or MCP access for Claude, ChatGPT and Cursor.
Why does the broken link checker miss some dead links?
By design. It only reports links it can prove are dead, because the fix a user applies to a false positive is deleting a link that was working. Anything it can’t positively classify is left out of the report rather than guessed at. It also checks internal links only — external link checking needs real HTTP requests to other people’s servers, which is a separate feature with its own throttling and failure modes.
Are the fan-out sub-queries the real ones Google ran?
No, and the panel says so. Query fan-out is probabilistic and Google doesn’t publish the sub-queries it generates, so nobody can show you the exact set — every tool in this space either models them or asks an LLM to guess. AuditAE models them from your own site: your focus keyword, your categories, your post titles, and your Organization schema. The wording won’t match Google’s, but the themes do, and that’s what decides whether a passage of yours can be retrieved. Nothing is sent off your server to produce them.
Why do reformulations like “what is X” never appear in the list?
Because they can’t be scored honestly. A reformulation restates your seed topic in different words, so there’s no distinguishing term to look for — any on-topic post covers it by construction. Including such a row would mean either always passing it (meaningless) or reporting a post that opens with a textbook definition as missing its own definition (wrong). The gaps that matter live in the other five fan-out types, so those are what the panel shows.
Which AI bots does the Crawler Tracker identify?
GPTBot, OAI-SearchBot, OAI-AdsBot, ChatGPT-User, ClaudeBot, Claude-User, Claude-SearchBot, Claude-Web, anthropic-ai, PerplexityBot, Perplexity-User, Bytespider, CCBot, cohere-ai, Diffbot, FacebookBot, Meta-ExternalAgent, Meta-ExternalFetcher, Bingbot, YouBot, Amazonbot, DuckAssistBot, PetalBot, TimpiBot, MistralAI-User, AI2Bot, iaskspider, AwarioRssBot and AwarioSmartBot. The list is curated to AI engines; we deliberately exclude Googlebot because that’s already covered by Google Search Console. Google-Extended and Applebot-Extended aren’t on it because they aren’t crawlers: they’re robots.txt tokens that Googlebot and Applebot read to decide whether your content may be used for AI, so no request ever carries them. The robots.txt audit still checks both.
Why does the crawler log show fewer hits than my server logs?
Because the plugin only sees requests that reach WordPress. A page cache (WP Rocket, LiteSpeed Cache, W3 Total Cache, WP Super Cache, your host’s cache, or a CDN such as Cloudflare) answers many requests before WordPress loads, and those hits are never logged. Treat the counts as a minimum: if a bot shows up at all, it is reading your site. For fuller counts, add the AI bots’ user agents to your cache’s list of user agents it never caches (WP Rocket: “Never Cache User Agent(s)”; LiteSpeed Cache: “Do Not Cache User Agents”; W3 Total Cache and WP Super Cache: “Rejected User Agents”): GPTBot , OAI-SearchBot , ChatGPT-User , ClaudeBot , Claude-User , Claude-SearchBot , PerplexityBot and Perplexity-User . A CDN needs the same exception as a cache-bypass rule on its side. Those bots’ requests then run through WordPress, which costs a little more server time.
Will the plugin slow down my site?
No. Crawler detection is a single substring scan of the User-Agent on template_redirect ; only a matching AI bot adds one row to a small indexed table, which is kept to 90 days and capped at 200,000 rows by a daily background clean-up, never inside a visitor’s request. A 404 that matches none of your redirects is answered without loading the redirect list. The /llms.txt response is cached in a transient and regenerated only when you save changes. Schema injection runs once per page render in wp_head . IndexNow pings are batched and sent from one background event. On any reasonable host, the per-request cost is well under a millisecond.
Where is data stored?
Locally, on your own site — no telemetry. Apart from IndexNow pings (changed URLs only; on by default, see External Services), nothing leaves your site unless you pair it with AuditAE. AI crawler hits live in one small custom table, {prefix}auditae_crawler_hits (pruned to 90 days and 200,000 rows, dropped on uninstall); everything else is in wp_options : auditae_crawler_rollup (per-bot first-seen + lifetime totals), auditae_redirects (redirect rules) and auditae_redirect_sources (an index of their source paths), auditae_audit_log (action log for the optional Connect tab), auditae_llms_txt_config (llms.txt manifest), auditae_organization_schema (site-wide schema), auditae_score_history / auditae_score_best (AI Readiness Score trend), auditae_score_email (the optional digest opt-in), auditae_link_index (the internal link graph), and IndexNow settings and its queue of changed URLs waiting to be sent — plus per-post meta _auditae_schema_graph and _auditae_faq_schema . On paired sites that used image compression, the original images AuditAE replaced are kept as files in wp-content/uploads/auditae-originals/ , recorded in _auditae_media_backup post meta. That folder carries a deny-all .htaccess ; nginx ignores .htaccess , so on nginx add location ^~ /wp-content/uploads/auditae-originals/ { deny all; } to your server block to keep it private. Uninstalling removes all of it except those originals, because they are the only copy of your images as they were: restore them first (Media Library → Restore original), or delete the folder yourself if you don’t need them.
What does the AuditAE connection cost?
Pairing is free. AuditAE itself is pay-per-use: AI-engine checks cost 5¢–12¢ each depending on the engine, WordPress edits cost 5¢ each (25¢ for a designed page; compressing images adds 1¢ per 5 images), and reads are free. Replies from AEBOT, the AuditAE assistant, are billed per reply on the AI tokens they use, and each reply shows its exact charge. New accounts get $5 of credit and there is no subscription. Sign up at https://auditae.app.
How do I revoke the AuditAE connection?
Settings → AuditAE → Connect to AuditAE tab → Disconnect. The bearer token is wiped instantly.
Does AuditAE conflict with Yoast SEO, Rank Math or other SEO plugins?
No — it’s designed to run alongside Yoast SEO, Rank Math, All in One SEO, SEOPress and The SEO Framework. AuditAE is an AI-search toolkit, not an SEO plugin: it never outputs title tags, meta descriptions, canonical URLs, Open Graph tags or sitemaps of its own, so it won’t trigger a “you have two SEO plugins active” warning. It changes values your SEO plugin stores only when asked: AEBOT can edit a post’s SEO title and description with your permission, and renaming an image updates the social-image URLs your SEO plugin saved, so share cards don’t break. Where the two genuinely overlap — structured data and IndexNow — AuditAE detects your SEO plugin and steps aside:
- Schema: it leaves the node types your SEO plugin owns (Organization, WebSite, Article, BreadcrumbList…) to that plugin and adds only the types it doesn’t cover, such as HowTo or Product, so you never get duplicate JSON-LD. The one exception: if your SEO plugin demonstrably emits no Organization or Person node and you’ve saved one in AuditAE, AuditAE emits yours so the site isn’t left without a publisher entity.
- FAQ schema: it skips posts that already carry an SEO plugin’s FAQ block, or Rank Math Schema Builder FAQ schema.
- IndexNow: it turns its auto-ping off if Rank Math (or a standalone IndexNow plugin) is already submitting.
You keep AuditAE’s AI-specific features — crawler tracking, AI Readiness Score, llms.txt — on top of your existing SEO setup, and a note on the plugin’s own settings page says exactly what’s been deferred.
How do I connect Claude, ChatGPT or Cursor to my WordPress site?
Pair the site on the Connect to AuditAE tab (a free AuditAE account is needed), then add the AuditAE MCP server, https://auditae.app/api/mcp , to your AI client. In Claude or ChatGPT, add it as a custom connector and sign in with your AuditAE account; that one connection covers every paired site in your workspace. In Claude Code, Cursor, Windsurf and other clients that take a config file, connect with an API key from your AuditAE dashboard; team workspaces connect with an API key too. The assistant can then read, draft, edit and publish posts, edit SEO titles and descriptions, and manage schema and llms.txt, limited to the permissions you granted on the Connect tab. MCP reads are free; edits are 5¢ each (25¢ for a designed page). Every change is listed on the Activity tab, and Disconnect revokes access at once.
What is IndexNow and do I need to set it up?
IndexNow is an open protocol that lets your site tell search engines “this URL just changed” instead of waiting for them to re-crawl on their own. The participating engines are Microsoft Bing, Yandex, Seznam.cz, Naver, Yep and Amazon, and one submission reaches all of them. Bing’s index also feeds AI answers such as Copilot’s, so faster indexing means faster eligibility to be cited. There’s nothing to set up — the plugin generates and serves the required ownership key automatically, and starts pinging on your next publish. It pauses on its own on staging and development sites and while Settings → Reading → “Discourage search engines from ind
