Free WordPress Plugin

AI Discovery Files
The free llms.txt & AI visibility plugin for WordPress

A free llms.txt plugin that does not stop at llms.txt. It generates all 10 AI Discovery Files, including ai.txt, identity.json and brand.txt, then logs which AI bots come and read them. Helps ChatGPT, Claude, Gemini and Perplexity discover, understand and correctly cite your website. Free under the GPL, with no pro tier and no feature gating. No coding required.

Download from WordPress.org Try Live Preview
AI Discovery Files WordPress plugin icon: generates llms.txt and 9 more AI visibility files

Most AI visibility plugins only generate llms.txt. This plugin generates all 10 files defined in the AI Discovery Files Specification, the complete set that AI systems need to accurately represent your business.

What Are AI Discovery Files?

AI Discovery Files are machine-readable files placed in your website's root directory that tell AI systems (ChatGPT, Claude, Gemini, Perplexity, and Copilot) exactly who you are, what services you offer, and how to refer to your brand. Think of them as robots.txt for AI, except instead of telling crawlers what to ignore, you tell AI systems what to get right.

Every one of the ten has a published specification behind it, with required fields and validation rules. Without them, AI systems fall back on guesswork: confusing businesses with competitors, inventing services, and getting brand names wrong. This plugin generates all ten from your WordPress dashboard, with no coding required.

  1. ADF-001llms.txtYour business identity in Markdown, the file most AI systems look for first.
  2. ADF-002llm.txtThe singular filename, redirected to llms.txt so either request resolves.
  3. ADF-003llms.htmlThe same identity as a readable web page, with Schema.org markup.
  4. ADF-004ai.txtHow AI systems may interact with, represent and cite your business.
  5. ADF-005ai.jsonThose same permissions as strict JSON, so nothing is left to interpretation.
  6. ADF-006identity.jsonCanonical facts about the business, aligned to Schema.org Organisation.
  7. ADF-007brand.txtHow your name is written, and what it must never be shortened to.
  8. ADF-008faq-ai.txtCitation-ready answers to the questions people actually ask about you.
  9. ADF-009developer-ai.txtTechnical context: your platform, your APIs and your integration points.
  10. ADF-010robots-ai.txtAccess rules for AI training and inference crawlers, bot by bot.

Why Your WordPress Site Needs AI Visibility

AI systems like ChatGPT, Claude, and Gemini are answering questions about your business right now. The question is whether they are getting it right.

Without AI Discovery Files

AI systems guess who you are based on whatever fragments they find. They confuse your business with competitors, fabricate services you do not offer, and cite your brand name incorrectly.

With AI Discovery Files

You give AI systems a single, authoritative source of truth: your name, services, location, permissions, brand rules, and FAQs. They stop guessing and start citing you accurately.

How to Add llms.txt to WordPress in 4 Steps

Four steps from install to AI-visible. No coding required.

1

Install

Search for "AI Discovery Files" in your WordPress plugins directory and click Install.

2

Configure

Fill in your business details. The plugin auto-detects your site name, tagline, pages, and more.

3

Preview

See exactly what each file contains before enabling it. Copy and inspect every line.

4

Go Live

Enable the files you want. They are served instantly at the correct root URLs.

What the AI Discovery Files Plugin Does

Everything a free llms.txt WordPress plugin should do, and the nine other AI Discovery Files besides.

Auto-Fill from Your Site

New in v2.0. Reads your own pages to pre-fill FAQs, services and key people as reviewable drafts. Needs WordPress 7.0.

Pulls your brand boilerplate, tagline and service descriptions from pages you already published, using the AI connectors added in WordPress 7.0. It copies rather than invents, and every import arrives as a draft you approve first.

Auto-Detection

Pulls your site name, tagline, pages, theme, and WordPress version automatically. Less typing, fewer mistakes.

Reads site options and post types rather than parsing your layout, so it behaves the same on classic themes, block themes and Gutenberg. On a standard install there is nothing to configure before the first file is generated.

All 10 File Types

Generates every AI Discovery File defined in the specification, from llms.txt to robots-ai.txt.

It is an llms.txt generator that does not stop at llms.txt. Most plugins produce that one file. All ten means your permissions, identity, brand rules and crawler policy agree with each other instead of one file speaking alone.

Live Preview

See exactly what each generated file contains before enabling it. Copy to clipboard with one click.

If you have been working out how to create llms.txt by hand, this is the shortcut: read the finished file, adjust the source fields, publish only when it says what you meant. Nothing is served until you switch it on.

Specification Validation

Checks every file against the AI Discovery Files specification and flags formatting or content issues.

A built-in validator for required fields, section headings and attribution, run against the published ADF-001 to ADF-010 specifications. For answer engine optimisation the parse matters: a malformed file is a file an AI system skips.

Conflict Detection

Warns if physical files already exist at the same URLs. No silent overwrites, no surprises.

If you already uploaded an llms.txt file to WordPress by hand, or a caching or SEO plugin is serving one, you are told before anything changes. Your existing file is never deleted and never overwritten.

Three Tiers

Start with Essential (2 files) and expand to Recommended (6) or Complete (all 10) when ready.

Essential is the fastest way to add llms.txt to WordPress and be done with it. Recommended adds permissions and identity. Complete publishes the full set. Move up whenever you like; nothing already written is rewritten.

AI Crawler Analytics

Logs which AI bots read your files, and when. Our own site recorded 43 reads from 11 bots in 30 days.

Named agents, not estimates: GPTBot, ClaudeBot, the Claude web crawler, PerplexityBot, OAI-SearchBot and the rest, with timestamps, response codes and CSV export. Being read is the step before being cited, and this is the part you can verify.

Built for Agencies

Multisite-aware with per-site settings, plus filters on every generated file and every collected data point.

Roll AI visibility out across a portfolio of client sites: per-site settings on a network, generation scriptable from WP-CLI, and the aidf_generated_content and aidf_collected_data filters to enforce your own house style.

No Lock-In

Disable the plugin and the files stop being served. No residual data, no side effects. Your content stays yours.

Free on WordPress.org with no pro tier, no subscription and no account to create. Every feature on this page is in the free download, the crawler log included.

Flagship feature

AI Crawler Analytics: Proof That AI Reads Your Site

Every other AI visibility tool asks an AI system what it thinks of you, then sells you the answer as data. This plugin does the one thing you can actually verify: it logs each AI crawler that fetches your AI Discovery Files, names the bot, timestamps the request, and records the response code. No prompts, no sampling, no guessing.

Every screenshot on this page is my own live dashboard on mcneece.com, covering the 30 days to 15 July 2026. Not a mockup, not demo data.

File reads
43

in 30 days

Distinct bots
11

named and counted

AI operators
8

OpenAI to ByteDance

Blocked
0

bots hit robots.txt

The live dashboard on mcneece.com. Four headline numbers, then a stacked daily chart colour-coded per bot, with 7, 30 and 90-day views.

Three things my own logs taught me

brand.txt

The files everyone skips are the ones bots read most

I expected llms.txt to dominate. It did not. brand.txt and faq-ai.txt tied on 8 reads each, ahead of llms.txt on 7. OAI-SearchBot's single most-requested file was /brand.txt: OpenAI's search crawler wanted to know what to call me before it wanted my page list.

5 → 8

The crawler mix changes underneath you

In a June window on this same site, 5 AI operators showed up. By this July window it was 8. ByteDance's Bytespider, PerplexityBot, GoogleOther and ClaudeBot all appeared for the first time. If you checked once in spring and concluded AI ignores you, your conclusion has expired.

29/29

Silent blocking is the real failure mode

Every logged request returned 200 and nothing was blocked. That sounds unremarkable until you learn how often it goes the other way: a stray robots.txt rule or a host-level bot filter turns AI crawlers away, and without a log you never find out.

Bot Breakdown table Every bot named, categorised, and counted. The Top File column is the one most dashboards miss: it shows OAI-SearchBot reaching for /brand.txt before anything else.
Discovery File Access File by file, which bots read what. llm.txt sits honestly at zero: the plugin reports what happened, not what you hoped happened.
Crawler Activity Log The raw evidence: 29 timestamped requests, filterable by bot, date, status code and file, exportable to CSV. Every row here returned a 200.
Crawler Logging Settings Off by default. You choose the retention period and tick exactly which of the 43 crawlers to track, grouped by operator.

What you get

Discovery File Access dashboard

Which bots read which files, with access bars, per-file bot tags, and 7, 30 or 90-day views. Logged at the moment the plugin serves the file, so the data holds up regardless of CDN or caching.

Works behind a CDN

Page-level analytics break behind edge caches, because the edge answers and your server never hears about it. This tracks discovery file responses the plugin serves directly, so it reports honestly on shared hosting, managed WordPress and edge-cached setups alike.

robots.txt conflict detection

Catches the contradiction where your robots.txt blocks an AI crawler while your permissions say "allow all", and flags it with a link to the fix. This is the check that turns "0 blocked" from a hope into a fact.

Drill-down and CSV export

Click any bot for its trend, file list and response codes. The activity log filters by bot, date range, status code and filename, and exports to CSV so the evidence leaves the dashboard with you.

What this does not tell you

A log line proving PerplexityBot fetched your faq-ai.txt is not proof that Perplexity quoted your answer. Retrieval and citation happen inside systems nobody outside those companies can observe, and any tool claiming otherwise is inferring from a sample and calling it measurement. This is deliberately inputs data: it verifies your files exist, are reachable, and are being fetched by named bots at recorded times. That is the part you control, and it is the part that is actually checkable. If you want the reasoning behind that split, read why we think AI mention trackers are a waste of money.

Logging is off by default, so there is zero overhead until you switch it on. It recognises 43 AI crawlers across 8 operator groups, including GPTBot, ClaudeBot, PerplexityBot, GrokBot, Applebot, Bytespider and Gemini-Deep-Research, and you choose which ones to track. Want the longer write-up? We published 11 days of raw crawler logs from this same site, and 30 days of dashboard evidence answering the sceptics who say nothing reads these files.

Essential, Recommended and Complete File Tiers

Start with the essentials and expand when you are ready. Every tier adds more signals for AI systems.

Essential

Start here: who you are and how AI can use your content.

Recommended

Improve personality, identity, permissions, brand control, and FAQs.

Complete

Full coverage with developer context, crawler directives, and compatibility.

Already Using Yoast or Rank Math for llms.txt?

Then you have a file at that address. Whether it tells an AI system anything useful about your business is a separate question.

Yoast SEO has generated llms.txt free since version 25.3, Rank Math and All in One SEO do the same, and between them they run on 17 million sites. So it is worth being precise about what they actually put in that file, because it is not what most people assume.

They generate a list of your content. Rank Math writes each post's title, URL and a short description, with options for post types and taxonomies. Yoast picks recent and cornerstone pages. What you get is an index of links: a sitemap, written in Markdown, for language models.

That is a fair reading of the original llms.txt proposal, which was written for documentation sites where a curated list of pages is exactly the useful thing. It is much less useful for a business. A list of forty blog post URLs does not tell ChatGPT what you do, which towns you cover, what you charge, or which jobs you turn down.

This plugin publishes a description of the company instead. Not a list of pages, and not limited to what happens to be written on your website. Most of the following appears nowhere on a normal site, because there has never been a reason to put it there:

That is the real difference. One plugin tells an AI which of your pages to read. This one tells it what your company is, who runs it, what it will and will not do, and how it expects to be quoted. Then it logs which AI systems came and took that information. Same filename, an entirely different job.

What each plugin actually writes into llms.txt, August 2026. Install counts as shown on WordPress.org.
Plugin Installs What its llms.txt contains Tells you which AI bots read them Cost
AI Discovery Files New A description of the company
company facts, key people, brand and citation rules, exclusions, plus 9 more files		 Yes, per file Free
Yoast SEO 10 million+ A list of pages
recent and cornerstone content		 No Freemium
Rank Math SEO 4 million+ A list of pages
title, URL and short description		 No Freemium
All in One SEO 3 million+ A list of pages
plus a full-text llms-full.txt		 No Freemium

You can run both. If your SEO plugin already writes a physical llms.txt, this plugin detects the clash and tells you, so you choose which one serves the file. Nothing breaks, and nothing is silently overwritten.

Want the evidence rather than the claim? We published 30 days of AI crawler logs showing which bots read which file on a live site.

Frequently Asked Questions

Everything you need to know about the AI Discovery Files WordPress plugin.

Getting Started

Installation, setup, and first steps with the plugin.

How do I install the AI Discovery Files plugin?

Install it directly from your WordPress dashboard. Go to Plugins → Add New, search for "AI Discovery Files", and click Install Now. Alternatively, download it from WordPress.org and upload the ZIP file.

What WordPress version do I need?

The plugin requires WordPress 7.0 or later and PHP 7.4 or later, and is tested up to WordPress 7.1. WordPress 7.0 matters because the optional AI import uses the AI Client that shipped in that release. Everything else works without it.

Is the plugin really free?

Yes, completely free and open source under the GPL v2 licence. There are no premium tiers, no upsells within the plugin, and no feature gating. Every capability is available to every user.

What is llms.txt and why does my WordPress site need it?

llms.txt is a plain-text file placed in your website root that tells large language models who you are, what you do, and what pages matter most. It is rapidly becoming the standard way websites communicate with AI systems, adopted by companies like Stripe, Cloudflare, and Dell. This plugin generates llms.txt plus 9 additional AI Discovery Files directly from your WordPress data, giving AI systems the complete picture rather than just one file.

How do I make my WordPress site visible to ChatGPT and other AI?

Install this plugin, fill in your business details, and enable the files. The plugin generates machine-readable files at URLs like yoursite.com/llms.txt and yoursite.com/ai.txt that ChatGPT, Claude, Gemini, and Perplexity read when they encounter your site. You can verify it is working with the free AI Visibility Checker.

How is this different from the paid AI Discovery Files Pack?

The AI Discovery Files Pack is a done-for-you service where experts research your business and write the files for you. The WordPress plugin lets you create and manage the files yourself through a settings interface. The plugin is for WordPress sites; the Pack works for any website.

Technical Questions

Caching, conflicts, performance, and developer features.

Does this work with caching plugins?

Yes. AI Discovery Files are served via WordPress rewrite rules with appropriate cache headers. They are compatible with WP Super Cache, W3 Total Cache, LiteSpeed Cache, and other popular caching plugins.

What if I already have a physical llms.txt file?

The plugin detects existing physical files and warns you on the Status tab. It will not override physical files. You should remove the physical file before enabling the plugin-generated version.

Can I customise the generated files?

Yes. Every generated file can be filtered using WordPress hooks. Use the aidf_generated_content filter to modify any file's output, or aidf_collected_data to adjust the data before generation.

What is the difference between llms.txt and robots.txt?

robots.txt tells search engine crawlers which pages to index or ignore. llms.txt tells AI systems who you are, what you do, and what pages contain the most important information. They serve different purposes and both should be present. This plugin generates llms.txt alongside robots-ai.txt, which handles AI-specific crawler permissions, giving you complete control over both discovery and access.

Will this slow down my website?

No. File generation only runs when specific URLs are requested (e.g. /llms.txt), so there is zero overhead on normal page loads. AI Crawler Analytics only tracks access to discovery file URLs that the plugin serves, so normal page requests are completely unaffected. When logging is disabled (the default), no tracking runs at all.

What is AI Crawler Analytics?

AI Crawler Analytics logs which AI bots read the AI Discovery Files this plugin generates: llms.txt, ai.txt, identity.json, and the rest. You see which bots accessed which files, how often, when they last called, and whether any are being blocked by your robots.txt. That gives you direct evidence that GPTBot, ClaudeBot, PerplexityBot and other AI crawlers are fetching your files, rather than an inference from prompting an AI system. On our own site it recorded 43 file reads from 11 distinct bots across 8 AI operators in the 30 days to 15 July 2026, and 55 reads from 9 bots in the 30 days to 4 August. We published the full log and what it does and does not prove. The feature works reliably on every hosting platform, including sites behind CDN edge caching, because the plugin controls the discovery file responses. It includes bot detail drill-downs, a filterable activity log with CSV export, and a WordPress dashboard widget. See the AI Crawler Analytics screenshots for what it looks like with real data.

What crawler data is stored, and is it GDPR compliant?

When AI Crawler Analytics is switched on, the plugin writes one row per request to a discovery file: the bot name matched from the user agent, which file was requested, the timestamp and the HTTP status code. Rows are stored in your own database on your own hosting. Nothing is sent to us or to any third party.

A daily task deletes anything past your retention window, which defaults to 90 days and cannot be set below 30. Logging is off by default, so no data is collected at all unless you opt in.

Because crawler logs can constitute personal data under UK GDPR, treat this as you would your server access logs: set a retention period you can justify, and mention the logging in your privacy policy. If you would rather not store it, leave the feature switched off. Every other part of the plugin works without it.

What happens if I uninstall the plugin?

Deactivating stops the files being served, so yoursite.com/llms.txt and the rest return 404 again. Nothing else on your site changes, because the plugin never writes files to disk and never edits your posts or pages.

Deleting the plugin removes its settings and its crawler log tables, so no orphaned data is left behind. Export the crawler log to CSV first if you want to keep the history.

Which AI crawlers does the plugin track?

The crawler registry covers 43 AI crawlers grouped into 8 categories: Major AI Assistants (GPTBot, ChatGPT-User, OAI-SearchBot, ClaudeBot, Claude-User, Claude-SearchBot, PerplexityBot, Perplexity-User, GrokBot, xAI-Grok, MistralAI-User), Google AI (GoogleOther, Google-CloudVertexBot, Gemini-Deep-Research, Google-NotebookLM), Apple and Microsoft (Applebot, Applebot-Extended, bingbot, AzureAI-SearchBot), Meta, Amazon, AI Search Engines, Training Crawlers (including CCBot and Bytespider), and Chinese AI (including DeepSeek and PanguBot). The default selection covers the major platforms, and you can tick or untick any individual crawler.

AI Visibility & SEO

How the plugin affects your visibility to AI systems and search engines.

Is this an AI SEO plugin or a GEO plugin?

AI Discovery Files focus on the infrastructure side of generative engine optimisation (GEO). Rather than tracking how often AI mentions your brand, this plugin ensures AI systems have accurate, structured data about your business to work with. It complements traditional SEO plugins like Rank Math or AIOSEO. They handle search engine optimisation, while AI Discovery Files handle AI visibility.

Which AI systems read these files?

AI Discovery Files are designed for all major AI systems including ChatGPT (OpenAI), Claude (Anthropic), Gemini (Google), Perplexity, Copilot (Microsoft), and other large language models. The llms.txt format in particular has been adopted by companies including Stripe, Cloudflare, and Dell.

How do I control what ChatGPT says about my business?

AI systems draw from whatever information they can find about your business. AI Discovery Files give you a way to provide authoritative, structured data: your correct business name via brand.txt, your actual services via identity.json, and pre-answered FAQs via faq-ai.txt. This does not guarantee what AI says, but it gives AI systems a reliable source to cite rather than guessing from third-party content.

What happens when I deactivate the plugin?

The files simply stop being served. Your settings are preserved so you can reactivate later without re-entering anything. If you delete the plugin entirely, all settings are removed cleanly.

How do I verify my files are working?

Use the free AI Visibility Checker to scan your site after enabling your files. It validates every AI Discovery File, checks for conflicts, and gives you a score with actionable recommendations.

Can I get listed in the AI Visibility Directory?

Yes. Once your files are live, submit your site to the AI Visibility Directory, the verified registry of websites implementing AI Discovery Files. It provides additional visibility and a dofollow backlink to your website.

Ready to Be Visible to AI?

Install the free plugin, configure your business details, and start being accurately represented by AI systems today.

Download from WordPress.org Check Your Site Free

Free under GPL v2 · No lock-in · No tracking

Changelog

v2.1.2
July 2026
  • AI Discovery Files load in a fraction of the time: These files must reach PHP on every request so crawler logging keeps working, which means they cannot be edge-cached. On sites with a heavy plugin stack, each request previously loaded the whole stack just to emit a small text file. The rendered files are now cached and served before page builders and other heavy plugins load, so llms.txt, ai.txt, ai.json and every other file respond quickly. Faster responses also let Google's PageSpeed and Lighthouse "Agentic Browsing" audit fetch llms.txt successfully instead of timing out.
  • Crawler analytics unchanged: Every request still reaches PHP and is logged. The cache rebuilds automatically when you save settings, edit a page or post, or update the plugin.
v2.1.1
June 2026
  • No more HTML entities in plain-text files: Quotes, ampersands, and similar characters in your settings now appear in llms.txt, ai.txt, brand.txt, faq-ai.txt, developer-ai.txt, and robots-ai.txt as real characters instead of " and &. llms.html keeps proper HTML escaping; the JSON files were never affected.
v2.1.0
June 2026
  • Language declarations everywhere: Every generated file now declares its language from your WordPress site language (Settings → General). Text files carry a Lang: header, ai.json and identity.json carry a language property, and llms.html keeps its lang attribute. BCP 47 format, zero setup.
  • ai.txt rebuilt to the specification's section structure: Proper Permissions ("AI systems MAY") and Restrictions ("AI systems MUST NOT") sections generated from your existing permission settings, so AI visibility checkers recognise the file's policy sections.
  • ai.json now validates against the published JSON Schema: $schema reference, permissions and restrictions as structured action lists, plus attribution, scope, licensing, and metadata. ai.txt and ai.json render from the same internal mapping, so they can never contradict each other.
  • identity.json rebuilt on Schema.org vocabulary: legalName, alternateName, sameAs, contactPoints, identifier, areaServed, a $schema reference, and the new Organisation Type field. Country names are converted to ISO codes for addressCountry.
  • robots-ai.txt in real robots syntax: User-agent: / Allow: / Disallow: stanzas per RFC 9309 plus a Sitemap: line, replacing the previous shorthand.
  • llms.html upgraded: Schema.org Organization JSON-LD, meta description, self-referencing canonical, mobile-friendly responsive layout, and indexable (index, follow).
  • Per-FAQ source links: An optional link URL on each FAQ emits the specification's URL: line in faq-ai.txt so AI systems can cite the page behind the answer.
  • Heads-up for integrators: The ai.json and identity.json output structures have changed. The old shapes did not validate against the published schemas; the new ones do.
v2.0.2
June 2026
  • Retired Claude-Web removed from robots-ai.txt: Anthropic no longer uses the Claude-Web user agent. The generated file now lists Anthropic's three current crawlers, each controlled independently: ClaudeBot (model training), Claude-User (live visits when a Claude user asks about your site), and Claude-SearchBot (search indexing).
  • Three more current crawlers added: OAI-SearchBot (OpenAI search indexing), Perplexity-User (Perplexity live visits), and Applebot-Extended (Apple AI training). Applebot-Extended follows your training policy in the same way as Google-Extended, CCBot, Bytespider, and Meta-ExternalAgent.
v2.0.1
May 2026
  • Minimum WordPress raised to 7.0: v2.0 is built around the AI Import feature which depends on WordPress 7.0's Abilities API, AI Client and Connectors API. Sites on WordPress 6.2 to 6.9 can continue to use v1.4.0, which still generates all 10 AI Discovery Files.
  • Real availability detection: v2.0.0 enabled the AI Import buttons on any WordPress 7.0 site because wp_get_connectors() returns core's connector definitions regardless of whether an API key is actually set. v2.0.1 asks the WP AI Client's own provider registry whether any provider is configured, so the buttons now reflect whether an AI call can actually succeed.
  • Fixed Connectors link: The "Settings → Connectors" link was pointing at a URL that does not exist in WordPress 7.0 and returned "Sorry, you are not allowed to access this page". Now points at the real core page.
  • Hard client-side gate: AI Import buttons refuse to open the import modal when no provider is configured, with a clear message instead of opening a useless dialogue.
  • Fast-fail server side: AJAX endpoints check for a configured provider as the very first thing after authentication, so a credential removed mid-session fails fast with a useful message.
  • Disabled-button styling: Disabled AI Import buttons now look visibly disabled (greyed out, cursor not-allowed). Previously the attribute was set but had no CSS, so it looked identical to enabled.
v2.0.0
May 2026
  • AI Import: Pre-fill FAQs, key people, services, brand boilerplate, and taglines from your own website using WordPress 7.0's new Abilities API + Connectors API. No more retyping content that already exists on your About / Team / FAQ pages.
  • Strict gate: The import button is enabled only when WordPress 7.0 is installed AND an AI provider (OpenAI, Anthropic, Google) is configured at Settings → Connectors. On older WordPress, the button explains what is needed. The rest of the plugin keeps working on WordPress 6.2+.
  • Review-first UX: Every extraction opens a modal with draft entries and checkboxes. Nothing is saved until you tick what you want and click the page's main Save Changes button. Existing entries are never replaced.
  • Same-site security: Imports can only fetch from your own site or another site in your WordPress network. Cross-domain URLs are rejected.
  • Long-page handling: Large source pages (over 12 KB) are processed in chunks and results are deduplicated.
  • Always-on progress indicator: An animated spinner with descriptive status text is visible during every AI call.
  • AI fingerprint filter: Every string the AI returns is post-processed to strip em dashes, banned marketing phrases, and flag rhetorical patterns for review.
  • Token cost: BYOK. Worst-case full session: under 10p on a reasonably-priced model.
v1.4.0
May 2026
  • Discoverability: The robots.txt block now lists every active AI Discovery File URL as a comment (e.g. # llms.txt: https://example.com/llms.txt) instead of as a custom directive. Parser-safe with every robots.txt implementation, including strict validators. AI tools that deliberately scan for AI Discovery hints still read the URLs out of the comments.
  • Admin: Status tab Discoverability Hints panel updated to reflect the comments-based block format.
v1.3.9
May 2026
  • Discoverability: Plugin now appends a labelled block to robots.txt listing every active AI Discovery File so automated tools can find them without knowing the canonical URLs
  • Discoverability: Plugin now emits a <link rel="alternate"> reference in the HTML <head> for each active discovery file
  • Admin: New "Discoverability Hints" panel on the Status tab. Detects physical robots.txt at the webroot and shows the exact block to copy in by hand when WordPress can't inject it automatically
v1.3.8
May 2026
  • Compatibility: Tested on WordPress 7.0. Broadened PHP support to PHP 7.4 through 8.5
  • Crawler logging: Bot detection now runs inside serve_file() so every discovery-file hit is logged, including on sites behind CDN edge caching
  • Crawler dashboard: Chart.js visualisations, bot-detail view, file-access widget, and log viewer
  • Security: XSS, auth, and data-integrity hardening in the crawler registry
v1.3.7
March 2026
  • Fix: Domain verification code saves correctly
v1.3.6
March 2026
  • Specification attribution enabled by default for new installs
v1.3.5
March 2026
  • Rebuild Summary processes full log history for accurate dashboard data
  • 90-day chart shows readable week dates instead of ISO week numbers
  • Distinct colour palette, so all bot colours are clearly distinguishable
  • Dashboard auto-refreshes after Rebuild Summary completes
v1.3.4
March 2026
  • Replace CSS charts with Chart.js for polished, responsive visualisations
  • Stacked bar chart with animated transitions, index tooltips, and proper scaling
  • Bot detail trend chart rendered with Chart.js
v1.3.3
March 2026
  • Rebuild Summary button in Crawler Logging Settings for immediate dashboard updates
  • Chart now shows full date range with proper bar sizing even with sparse data
v1.3.1
March 2026
  • Move bot detection into discovery file server for reliable CDN/cache compatibility
  • Set s-maxage=0 on discovery file responses so CDN edge caches pass through to origin
  • Remove page-level tracking, so all analytics are now scoped to discovery file access only
  • Consolidate to single dashboard widget (Discovery File Access)
  • Update all UI labels and descriptions for file-access-only scope
v1.3.0
March 2026
  • New: AI Crawler Analytics, to see which AI bots access your AI Discovery Files
  • Discovery File Access dashboard with bot breakdown and visual access bars
  • robots.txt conflict detection with actionable alerts
  • Warning when WordPress "Discourage search engines" setting blocks all AI crawlers
  • Bot detail drill-down with access trends and file breakdown
  • Filterable activity log viewer with CSV export
  • WordPress dashboard widget (Discovery File Access)
  • 43 AI crawler definitions across 8 operator groups (GPTBot, ClaudeBot, PerplexityBot, GrokBot, Applebot, and more)
  • User controls: enable/disable, data retention (30-365 days), per-bot selection
v1.2.1
March 2026
  • Replace static page picker dropdowns with AJAX-powered search autocomplete
  • Page pickers now scale to sites with hundreds or thousands of pages
  • Products/Pricing field also searches WooCommerce product pages
  • Full keyboard navigation (arrow keys, Enter, Escape) and ARIA accessibility
  • Increase page limit in generated AI discovery files from 50 to 200
  • Fix Plugin Check errors: add translators comment and prefix global variable
v1.2.0
March 2026
  • Add review prompt banner on settings page with interactive star rating
  • Smart trigger: only shows after settings saved, 2+ active files, and 3+ days active
  • Stars 4-5 link to WordPress.org review page; stars 1-3 show support link
  • "Already reviewed" permanently dismisses; "Not now" snoozes for 90 days
  • Polished SVG star animations with brand-consistent design
v1.1.1
March 2026
  • Fix verification code input to accept just the token (auto-strips prefix if pasted)
  • Serve only the bare token in ai-visibility-verify.txt
v1.1.0
March 2026
  • Add domain verification for the AI Visibility Directory
  • New verification modal with step-by-step instructions
  • Serve ai-visibility-verify.txt with user-supplied verification code
  • Verify Domain button in directory CTA banner
  • Verification panel on Status tab
  • AJAX save for verification code (no page reload)
v1.0.0
March 2026
  • Initial release
  • Generate all 10 AI Discovery File types
  • Settings page with Identity, Permissions, Content, Technical, Preview, and Status tabs
  • Live file preview with copy-to-clipboard
  • File validation against specification
  • Conflict detection for existing physical files
  • Tier progress indicator (Essential, Recommended, Complete)
  • FAQ repeater for building faq-ai.txt
  • Developer hooks and filters for customisation