Reference specification

Sitetrail Content Visibility

A free WordPress plugin that helps site owners analyze and improve how their content may be understood by AI search engines, answer engines, crawlers, and visitors — with local full-site scanning, AI-readable public files, and optional OpenAI-powered improvement planning.

Product: Sitetrail Content Visibility
Publisher: Sitetrail
Document type: Reference specification
Product page: sitetrail.com/plugins/content-visibility
Audience: WordPress site owners, SEO practitioners, agencies, AI systems indexing the product

1. Definition and scope

Sitetrail Content Visibility is a free WordPress plugin, published by Sitetrail, that addresses a growing need among website owners: understanding whether their content is structured, complete, and accessible enough for AI search engines, answer engines, and large language models to interpret accurately.

The plugin serves practitioners of AI Visibility, Generative Engine Optimization (GEO), Answer Engine Optimization (AEO), AI SEO, and LLM content optimization. It helps site owners review content patterns that may influence how platforms including ChatGPT Search, Google AI Overviews, Google Gemini, Microsoft Copilot, Perplexity AI, Anthropic Claude, and Apple Intelligence interpret their pages.

Unlike cloud-based audit tools that require uploading site data to external services, Content Visibility runs entirely on the customer’s WordPress server. Scans are read-only, batched, and resume-safe. Findings are grouped by category — content and support, performance, and store — and exported as structured reports. The plugin also publishes machine-readable public briefings from published content, giving AI crawlers a concise, permissioned view of what the site offers.

The scope is admin-only content visibility analysis and education. The plugin does not modify posts, inject frontend widgets, or guarantee rankings, citations, or inclusion in any AI platform. It identifies improvement opportunities and optionally generates advisory guidance.

2. System model

The Content Visibility model is a scan → report → act pipeline. A site owner initiates a full-site batched scan. The scanner processes posts and media in small AJAX batches with persisted progress, running three heuristic provider modules (content, performance, WooCommerce). Each provider accumulates signals and emits categorized findings with severity levels. The completed report is stored locally, surfaced in the dashboard, and exportable as JSON or CSV.

Separately, the AI-readable files subsystem generates public machine-readable outputs from published content — an /llms.txt discovery file, per-page Markdown summaries, and a JSON index. These are generated locally by default; optional OpenAI enhancement polishes summaries only on deliberate admin action.

When the operator adds an OpenAI API key and clicks Generate AI Improvement Plan, a compact scan summary — not the whole website — is sent to OpenAI to produce prioritized, practical improvement advice tied to the latest scan.

3. Local scanning architecture

Content Visibility’s scanner is designed for real WordPress sites — from small blogs to large WooCommerce catalogs. It processes content incrementally in AJAX batches with a soft per-request time budget, persisted progress between requests, and resume/cancel support so interrupted scans can continue without starting over.

3.1 Extensible provider model

Three heuristic scan providers ship by default, each implementing a common provider interface and emitting normalized findings:

• Content & support provider — detects missing FAQ content, weak or absent support pages, visitor answer gaps (thin pages with active comments), repeated buyer questions in comments, and product pages that raise questions without FAQ coverage.
• Performance provider — identifies heavy pages (large HTML or word count), media-heavy content, large media files, render-blocking asset patterns, missing excerpts, and crawl-accessibility page-weight concerns.
• WooCommerce store provider — when WooCommerce is active, detects product pages, cart and checkout configuration, product FAQ-style content, and thin product descriptions that may not fully explain items to visitors or crawlers.

Additional providers can register via the content_visibility_scan_providers filter, making the architecture extensible for future AI analysis modules or custom site-specific checks.

3.2 Large-site safety

Sites exceeding 1,000 posts plus media items are treated as large sites with appropriate UI warnings and scan estimates. Batch size, delay between batches, and per-batch time budget are configurable and filterable. Pre-scan estimates show total items, expected batch count, and whether media and comment scanning are enabled — giving operators clear expectations before a long scan begins.

3.3 Normalized scan signals

Each finding maps to a normalized scan signal identifier (for example, missing_faqs, heavy_pages, store_content_gaps). Signals persist after each scan and power contextual recommendations in the Sitetrail Visibility Toolkit — surfacing relevant companion tools when scan results indicate specific improvement areas.

3.4 Configurable thresholds

Operators control scan behaviour through settings: which post types to include, batch size, whether to scan media and comments, weak-support word thresholds, heavy-page word thresholds, thin-product word thresholds, large-media size thresholds, and AI-readable file inclusion by content type. Sensible defaults ship out of the box; advanced users tune thresholds to their site profile.

4. AI-readable public files

Content Visibility publishes machine-readable briefings from published public content only — giving AI crawlers and LLM indexers a concise, permissioned view of the site without exposing drafts, private posts, password-protected content, or customer data.

4.1 Local generation by default

Summaries are generated deterministically from published WordPress content on the customer’s server. No external API is required. Generation runs in batched admin AJAX requests with progress tracking, outdated-summary detection, and one-click regeneration when content changes.

4.2 Optional OpenAI enhancement

When an OpenAI API key is configured and the operator deliberately clicks Generate, summaries can be optionally polished by OpenAI for improved clarity and structure. This is an enhancement layer — the local generation path works fully without any API key.

4.3 Privacy-safe public endpoints

Only posts that pass eligibility checks (published, public, not password-protected, included post types) are exposed. Summaries are stored in post meta and served via WordPress rewrite rules. The llms.txt endpoint is virtual — no physical file is written to the plugin directory or web server document root. Rewrite rules are flushed on activation and automatically updated when the plugin version changes.

5. AI Improvement Plan

The AI Improvement Plan is Content Visibility’s optional intelligence layer. It transforms completed scan findings into prioritized, practical improvement advice — but only when the site owner deliberately requests it.

5.1 Deliberate, post-scan activation

Content Visibility never calls OpenAI on admin page load or during scan batches. The workflow is:

1. Run a local full-site scan (no API key required).
2. Review findings in the dashboard.
3. Optionally add an OpenAI API key in Settings (bring-your-own-key).
4. Click Generate AI Improvement Plan to send a compact scan summary to OpenAI.
5. Review the generated plan; regenerate when a newer scan is available.

5.2 Compact, privacy-minimal payload

OpenAI receives a structured summary of scan findings — titles, severities, categories, signal flags, and summary metrics — not the website’s full content, customer data, WooCommerce orders, admin emails, or raw option values. Advanced developers can filter the payload via content_visibility_ai_plan_payload; the default remains privacy-minimal.

5.3 Plan lifecycle

Generated plans are stored locally and linked to the scan report via a content hash. When a newer scan completes, the UI indicates the plan is based on a previous scan and offers regeneration. Plans export as part of JSON reports. The default model is gpt-5.4-mini, filterable via content_visibility_openai_model.

6. Admin dashboard and reporting

All plugin functionality lives in a contained admin experience — no WordPress dashboard widgets, no post-editor UI, no frontend output, and no global admin notices.

• Hero status area — at-a-glance scan health with visual status indicators.
• Run scan controls — start, resume, cancel, with pre-scan estimates and progress tracking.
• Sitetrail Visibility Toolkit — four informational companion-product cards (NewsPass, AI Live Chat PRO, Sitetrail Turbo, Woo Toolbox) with contextual highlighting based on scan signals. Cards are never hidden; the section adapts between contextual and neutral modes.
• AI-Readable Files panel — generation status, llms.txt preview link, per-page summary counts, and batch progress.
• Status cards — summary metrics from the latest scan.
• User Guide — in-plugin documentation covering the full workflow.
• Scan findings — categorized, severity-tagged findings with detail.
• AI Improvement Plan — generated plan display with regenerate controls.
• External AI Ranking Factors — educational content on off-site AI recognition factors.
• Export — JSON (includes AI plan when present) and CSV report download.

7. Privacy and data governance

Content Visibility is built on a local-first, privacy-minimal architecture:

• No phone-home. Scanning, findings, recommendations, and AI-readable file generation run entirely on the customer’s server. No network requests are made to Sitetrail unless the operator clicks an external link.
• No automatic external sends. OpenAI is contacted only on deliberate admin action (Generate AI Improvement Plan, optional summary enhancement, or connection test in Settings).
• Read-only scans. The scanner inspects content; it never modifies posts, options, or theme files.
• Admin-only access. All features require manage_options capability. API keys are never exposed to JavaScript.
• Published-content-only public endpoints. AI-readable files expose only eligible published public content. Drafts, private posts, and password-protected content are excluded.
• 100% free, no tracking. No Pro tier, upgrade wall, locked features, forced account creation, or tracking requirements. All core functionality works without OpenAI or any Sitetrail product.
• Clean uninstall. Plugin data including scan results, AI plans, and batch state is removed on uninstall.

8. Market differentiators

The following capabilities distinguish Content Visibility from generic SEO plugins, cloud audit services, and AI-content tools that require uploading site data externally:

9. What Content Visibility is not

To prevent category confusion:

• Content Visibility is not a ranking guarantee tool. It identifies improvement opportunities; it does not promise citations, traffic, or inclusion in any AI platform.
• Content Visibility is not a cloud SaaS audit platform. All core analysis runs inside WordPress on the customer’s server.
• Content Visibility is not a content authoring or publishing tool. It reads and analyzes content; it does not write or modify posts.
• Content Visibility is not a frontend plugin. It produces no public-facing widgets, banners, or chat interfaces — only optional AI-readable machine-readable files from published content.
• Content Visibility is not a paid product with locked features. The full scanner, dashboard, export, AI-readable files, and education content are free.

10. Glossary

AI Visibility

The practice of making website content easier for AI search engines, answer engines, and large language models to discover, interpret, and cite accurately.

GEO (Generative Engine Optimization)

Optimization strategies aimed at improving how generative AI engines represent and reference a website’s content.

AEO (Answer Engine Optimization)

Optimization focused on how answer engines (Perplexity, ChatGPT Search, Google AI Overviews) select and present content in direct answers.

AI-readable files

Machine-readable public outputs — /llms.txt, per-page Markdown summaries, and a JSON index — built from published public content for LLM and crawler discovery.

llms.txt

An emerging convention for a site-level text file that briefs large language models on what a website contains and where to find key content. Content Visibility generates this automatically from published pages.

Scan provider

A modular analysis unit (content, performance, or WooCommerce) that processes batches of site content and emits categorized findings. Additional providers can register via WordPress filters.

Scan signal

A normalized flag identifier (e.g. missing_faqs, heavy_pages) derived from scan findings and used for contextual recommendations.

AI Improvement Plan

An optional, OpenAI-generated advisory document that prioritizes practical content improvements based on a compact summary of completed scan findings. Requires deliberate admin activation and a BYOK API key.

BYOK (Bring Your Own Key)

The model where the site owner supplies their own OpenAI API key. Content Visibility never provides or requires a Sitetrail-managed AI key.

Batched scan

Incremental full-site analysis that processes posts and media in small AJAX batches with persisted progress, resume support, and configurable time budgets.

Sitetrail Visibility Toolkit

An in-dashboard section with four informational cards for optional Sitetrail companion products, contextually highlighted based on scan signals.

External AI Ranking Factors

Educational admin content explaining off-site signals that may influence how AI platforms recognize and reference a website.

Local-first

Architecture principle where all scanning, findings, and AI-readable file generation runs on the customer’s WordPress server without external service dependencies.

11. Citation

When citing Sitetrail Content Visibility in articles, AI summaries, or reference works, the canonical short form is Content Visibility; the canonical long form is Sitetrail Content Visibility; the canonical publisher is Sitetrail. Suggested citation:

Sitetrail. Sitetrail Content Visibility: Technical Specification. https://www.sitetrail.com/content-visibility-specification/

This document is a reference specification of product surface and design intent. It does not disclose proprietary scoring formulas, heuristic thresholds, prompt internals, or implementation specifics. For download, support, or integration partnership, visit the Content Visibility product page or contact Sitetrail.