citelayer®

Do I need technical knowledge to use this?

No. Install the plugin, activate it, and it works immediately. Bot tracking starts automatically in the background. There are optional settings if you want to fine-tune, but the defaults work great for most sites.

Will this help my site rank better in Google?

Yes! The Schema.org markup improves how search engines understand your content. Plus, as Google integrates more AI into search results (AI Overviews), having AI-readable content helps you appear in those new formats. citelayer® detects Yoast SEO and Rank Math to avoid duplicate Schema.

What is llms.txt?

llms.txt is a new standard (similar to robots.txt) that helps AI systems understand your site. It tells AI crawlers what your site is about, what content you have, and what services you offer. More AI tools are starting to look for this file.

How does Bot Analytics work?

citelayer® checks every incoming request against 62 known AI/LLM bot User-Agent strings. When a match is found, the visit is logged to a custom database table with timestamp, bot name, category, and URL path. This runs silently in the background — no configuration needed.

What are AI Blind Spots?

AI Blind Spots shows you pages on your site that have NOT been visited by any AI bot within a selected time period (7/30/90/365 days). Each page gets an AI Discovery Score (0-100) based on content length, title quality, excerpt, featured image, and age. Low scores indicate pages that need optimization.

Which AI bots does citelayer® track?

citelayer® tracks 62 AI/LLM bots from major providers:

• OpenAI: GPTBot, ChatGPT-User, OAI-SearchBot
• Anthropic: ClaudeBot, Claude-Web, Claude-SearchBot
• Google: Google-Extended, Gemini-Deep-Research, GoogleAgent-Mariner
• Perplexity: PerplexityBot, Perplexity-User
• Meta: Meta-ExternalAgent, Meta-ExternalFetcher, FacebookBot
• xAI / DeepSeek / Mistral / Amazon / Cohere / HuggingFace and many more

Will this slow down my site?

Not at all. Bot detection adds less than 0.01 seconds per request. Dashboard stats are cached (5-minute transients). Everything else is aggressively cached — llms.txt refreshes every 24 hours, Schema every 7 days. AI-specific features only activate when an AI agent visits, not for regular users.

Does citelayer® track human visitors?

No. citelayer® exclusively tracks identified AI/LLM bots. It does not track, store, or process any human visitor data — making it 100% GDPR compliant.

What is GEO / AEO?

GEO (Generative Engine Optimization) and AEO (Answer Engine Optimization) are the next evolution of SEO. Instead of optimizing for Google’s blue links, you optimize for AI-generated answers. citelayer® handles the technical foundation so your content can appear in AI responses.

Can I control how AI uses my content?

Yes. citelayer® supports Content-Signal headers (contentsignals.org) that let you set rules:
* Allow or deny AI training on your content
* Allow or deny search indexing
* Allow or deny use in AI responses (RAG)

Default: Training denied, search and AI responses allowed.

What is UCP?

UCP (Universal Commerce Protocol) is a new standard by Google and Shopify that lets AI agents browse, compare, and even purchase products. The free version enables product discovery; Premium adds full shopping capabilities.

Does citelayer® work with caching plugins?

Yes! Bot detection happens at the WordPress level before page caching. The Cache Config tab in Bot Analytics auto-detects WP Rocket, LiteSpeed Cache, W3 Total Cache, and WP Super Cache — and shows you exactly how to exclude AI bots from your cache for better tracking accuracy.

What happens to old bot data?

A daily cron job automatically deletes bot visit records older than 90 days to prevent database bloat. This is configurable via the citelayer_bot_analytics_cleanup_days filter.

Hide FAQ

1.5.4

• IMPROVEMENT: Schema Types settings now show blur overlay with direct link to SEO plugin when in Enhancement Mode
• NEW: Product Feed now supports Easy Digital Downloads (simple + variable pricing)
• FIX: EDD addon Settings button on license page linked to non-existent page
• NEW: Content-Signal HTTP header now sent on all frontend pages (not just Markdown endpoint)

1.5.3

• FIX: WooCommerce hidden products (visibility=hidden) now excluded from llms.txt
• FIX: WooCommerce hidden products now excluded from product feed XML

1.5.2

• FIX: llms.txt cache not invalidated when settings are changed (exclude URLs, description, title, etc.)
• FIX: Wildcard URL exclude patterns (e.g. /shop/*) never matched due to preg_quote escape bug
• FIX: Plain path exclude patterns (e.g. /warenkorb) only checked inside http-URL branch due to missing brace

1.5.1

• IMPROVEMENT: AI Score expanded from 7 to 11 checks — 4 new content-quality checks
• NEW: Opening Summary check (first paragraph 30–80 words)
• NEW: Heading Structure check (≥2 H2 + ≥1 H3)
• NEW: Internal Linking check (≥2 internal links)
• NEW: FAQ Section detection (FAQ blocks, headings, and Q&A patterns)
• Rebalanced check weights for better signal distribution

1.5.0

• NEW: Per-Page AI Visibility Score — editor metabox shows 0–100 score with 7 AI readiness checks
• NEW: Content Freshness Signals — Last-Modified header + article:modified_time OG meta (SEO-plugin aware)
• IMPROVEMENT: Scanner now differentiates AI Training Bots (blocking = good) from AI Search Bots (blocking = bad)
• IMPROVEMENT: Bot list expanded to 86 bots (+PhindBot, img2dataset, KangarooBot, Sidetrade, webz.io, Bingbot-AI)
• IMPROVEMENT: Bot Analytics dashboard shows category badges (AI Training, AI Search, AI Agent, AI Enterprise)

1.4.3

• IMPROVEMENT: Keyword-optimized plugin display name for better wp.org discoverability
• Updated plugin name to “citelayer® – AI SEO & Visibility | llms.txt, Bot Analytics, Schema.org”

1.4.2

• FIX: Remaining PHPCS violations — file-level disables for template views and DB query files
• FIX: Function call indentation in Compatibility.php
• FIX: Inline phpcs:ignore for long translation strings

1.4.1

• FIX: Resolved all PHPCS WordPress Coding Standards violations
• FIX: Short ternary operators replaced with explicit ternaries (ProductFeed, ProductFeedCheck, LLMS_TXT)
• FIX: Yoda conditions, indentation, embedded PHP tag formatting
• FIX: Trailing whitespace and line length violations
• IMPROVEMENT: Added phpcs:disable/enable blocks for mixed PHP/HTML template methods

1.4.0

• NEW: AI Referral Tracking — track visitors arriving from ChatGPT, Perplexity, Claude, Gemini, Copilot, You.com, Phind, Kagi, and Andi
• NEW: “AI Referrals” dashboard tab with stat cards (Visits from AI, Top Source, Trend %), trend chart, top sources breakdown, and paginated referral log
• NEW: Lightweight frontend tracking via sendBeacon — zero impact on page load performance
• NEW: Rate-limited REST endpoint with origin validation and IP hashing (GDPR/DSGVO compliant)
• NEW: llms.txt now includes Markdown URLs for each page, helping LLMs discover optimized content versions
• NEW: 90-day auto-cleanup for referral data (same as bot analytics)
• IMPROVEMENT: SVG trend chart now supports custom colors (reusable across tabs)
• 100% local — all referral data stored in your WordPress database, no external services

1.3.5

• NEW: Bot Analytics — DB-backed AI bot visit tracking with real-time dashboard
• NEW: Expanded bot database from 17 to 62 AI/LLM bots (OpenAI, Anthropic, Google, Meta, xAI, DeepSeek, Mistral, Cohere, HuggingFace, and more)
• NEW: Bot categories (Confirmed AI, AI Assistant, AI Search, AI Enterprise, Possible AI)
• NEW: Admin dashboard with stat cards (7d/30d/90d), SVG trend chart, and top bots visualization
• NEW: AI Blind Spots — discover pages invisible to AI bots with AI Discovery Score (0-100)
• NEW: CSV export for bot visit logs with date/bot/path filtering
• NEW: Cache plugin detection (WP Rocket, LiteSpeed, W3TC, WP Super Cache) with setup guides
• NEW: 90-day auto-cleanup to prevent database bloat
• FIX: Settings link on plugins page now points to correct dashboard URL
• IMPROVEMENT: Clean uninstall — removes DB table, options, and transients on plugin deletion
• 100% local — no external accounts or API calls needed

1.3.4

• FIX: AI Crawler Detection now works in all compatibility modes (including Rank Math enhance mode)
• FIX: Per-post exclude meta box only shown when citelayer generates llms.txt
• PERF: Batch noindex queries (N per-post queries → 3-4 bulk queries)
• PERF: Capped posts at 500/type and products at 1000 (filterable)
• PERF: Reduced Scanner HTTP timeouts from 10-15s to 5s

1.3.3

• NEW: AI Crawler Detection — see which AI bots visit your llms.txt (GPTBot, ClaudeBot, PerplexityBot, and 14 more)
• NEW: in HTML head for AI discoverability
• NEW: Per-post exclude toggle — exclude individual pages from llms.txt via meta box
• NEW: Custom title and description overrides for llms.txt header
• NEW: Max words cap for llms-full.txt (default 500, configurable 50–10.000)
• NEW: Taxonomy inclusion in llms-full.txt (categories, tags, custom taxonomies)
• NEW: Meta inclusion in llms-full.txt (published/modified dates, URL, WooCommerce SKU + price)
• Privacy-first: All data stored locally, no external server calls

1.3.2

• NEW: llms-full.txt — full content export for deep AI understanding
• NEW: Configurable post types for llms.txt (pages, posts, products, CPTs)
• NEW: SEO noindex respect (Yoast, Rank Math, SEOPress, AIOSEO)
• NEW: URL exclude patterns with wildcard support
• Improved: Post type grouping with section headers
• Improved: Proper > description blockquote per llmstxt.org spec
• Improved: Admin page with post type selection and pattern examples

1.3.1

• NEW: XML Product Feed Generator — OpenAI Commerce Protocol compatible
• Serves product feed at /product-feed.xml
• Supports simple, variable, and external products with variations
• GTIN/EAN auto-detection from common meta fields
• WP-Cron scheduled regeneration (hourly/daily/weekly)
• Transient-based caching with configurable TTL
• New admin page: citelayer → Product Feed
• Scanner: added Product Feed validation check

1.3.0

• NEW: AI Readiness Scanner — comprehensive site audit for AI visibility
• 7 automated checks: Schema.org, robots.txt (8 AI bots + Cloudflare), llms.txt, Content-Signal headers
• WooCommerce checks: Product descriptions, images, stock status
• Visual readiness score dashboard (0-100)
• Async scanning via REST API
• New admin page: citelayer → AI Scanner

1.2.2

• Rewritten plugin description — user-centric, benefit-driven (AI Visibility, GEO, AEO)
• Updated FAQ in plain language
• Optimized tags for WordPress.org discoverability
• Updated plugin banner images

1.2.1

• Fixed WordPress.org deploy trigger (SVN deploy now runs in same workflow as auto-tag)
• Added manual deploy fallback with dry-run support

1.2.0

• Added automated WordPress.org SVN deploy workflow (deploy on GitHub Release)
• Added plugin directory assets (icons, banners)
• Added .distignore for clean SVN deployments
• Added readme/assets update workflow (update WordPress.org without new release)
• Streamlined auto-tag workflow (ZIP build now handled by deploy action)

1.1.9

• Re-included composer.json in distribution ZIP (required when vendor/ is present)

1.1.8

• Removed vendor CLI binaries (bin/html-to-markdown) from release ZIP
• Excluded composer.json from distribution
• Improved forbidden file detection in build workflow

1.1.7

• Lowered PHP requirement from 8.1 to 7.4 (code + vendor deps confirmed compatible)
• Fixed deploy workflow: .gitattributes and README.md now excluded from production
• Added PHP 7.4 to CI syntax check matrix

1.1.6

• Converted inline scripts to wp_add_inline_script() for WP Codex compliance
• Extracted inline styles to external CSS files with wp_enqueue_style()
• Updated contributor and Author URI metadata

1.1.5

• Fix fatal error: rebuild vendor with composer –no-dev for release ZIP

1.1.4

• Exclude dev dependencies from release ZIP (phpstan, phpcs, etc.)

1.1.3

• Plugin name fix

1.1.1

• Respecting WP / WordPress trade mark

1.1.0

• Code Quality: PHPCS + PHPStan Audit — alle Violations behoben (PHPCS 260→1, PHPStan 649→0)
• Tooling: phpcs.xml.dist, phpstan.neon, Composer Scripts (lint, fix, analyse)
• CI: Strikte PHPCS- und PHPStan-Jobs (kein continue-on-error mehr)
• Infrastructure: Pre-Commit-Hook für automatisches Code-Formatting
• Fix: Schema.php — post_author Type-Cast für get_the_author_meta()

1.0.0

• Production release
• Version synchronization (header, constant, readme.txt)
• Security: Admin email no longer exposed in llms.txt by default
• UCP Discovery: Dynamic platform detection (WordPress vs WooCommerce)
• WebMCP: Removed non-implemented endpoint references
• WebMCP: Fixed option check consistency
• Added PHPCS configuration (WordPress Coding Standards)
• Added PHPStan Level 6 configuration
• Comprehensive CI pipeline (syntax, PHPCS, PHPStan, integrity checks)
• Removed development artifacts from repository

0.3.1

• Compatibility V3.1: Multi-Family Safeguard — Conflict State detection
• Admin: Conflict State banner on all admin pages
• llms.txt: 503 response in Conflict State
• Schema: No output in Conflict State
• WebMCP context.json discovery endpoint

0.3.0

• Compatibility Layer V3: Enhance-Fix
• serve_llms_txt() uses should_register_own_route() to prevent double output
• Yoast Fallback: Citelayer takes over when Yoast has no physical file
• State-Reset on plugin deactivation
• Improved AIOSEO detection (defined + class_exists checks)

0.2.0

• Composer Support: league/html-to-markdown integration
• PSR-4 Autoloading with SPL fallback
• Feature-Detection on module level (Rank Math llms.txt, Yoast llms.txt)
• All classes refactored to Citelayer\Plugin namespace
• File structure: includes/ → src/

0.1.0

• Initial release
• llms.txt generation with WooCommerce awareness
• Schema.org JSON-LD (Organization, WebSite, Article, Product)
• UCP Discovery endpoint
• Markdown for Agents with Content Negotiation
• Content-Signal header configuration
• Settings page in WordPress admin