The All-in-One SEO, Speed & AI Suite for WordPress
Seven Plugins. One Powerful Suite.
Skill WP SEO Pro
- ✓Smart Meta Box (Gutenberg, Classic Editor & Elementor)
- ✓AI SEO Content Optimizer with 18-Point Live Scoring
- ✓AI Writer, Editor Assistants & Brand Voice
- ✓AI Keyword Generator & Deep Keyword Research
- ✓XML, News, Image & Video Sitemaps and JSON-LD Schema
- ✓GSC + GA4 Dashboard, Rank Tracker, PageSpeed & Multilingual
Skill WP SEO Tools Pro
- ✓Redirection Manager with Regex, Geo & AI 404 Suggestions
- ✓404 Monitor, Smart 404 Page & Broken Link Checker
- ✓Bulk Editor, Importer & Permalink Tools
- ✓Robots.txt, .htaccess Editor, HTML Sitemap & Schema Builder
- ✓Database Backup, Website Files Backup, Theme Cleanup & User Role Manager
Skill WP SEO Media Pro
- ✓WebP & AVIF Converter with Lossless & Background Queue
- ✓AI Image Alt Text, Title, Caption & Bulk Multilingual Generation
- ✓Video SEO — Schema, Sitemap, YouTube Facade & Lazy Loading
- ✓CDN Integration, Lazy Loading & Image SEO Best Practices
Skill WP Optimize Pro
- ✓Page Cache — Disk, Per-Device, Per-Role & Preload
- ✓Critical CSS, Minification, Combination & Unused CSS Removal
- ✓Database Optimizer, Heartbeat Control & Cron Optimization
- ✓Core Web Vitals Real-User Monitoring & CWV Recommendations Engine
Skill WP LLMs Gen Pro
- ✓Auto-Generated llms.txt with Granular Post-Type Control
- ✓llms-full.txt — Full Markdown Content Export & Per-Post Control
Skill GEO/AEO AI Pro
- ✓AI Overview, SGE & Answer Engine Citation Tracker
- ✓Entity Graph Builder, Geo Variants & Brand Voice Trainer
SKILLSEOLIC
- ✓License Key Generation, Domain Locking & Multi-Domain Plans
- ✓Enhanced Dashboard (v7.2.9), Plugin Status Grid & REST API Reference
Skill WP SEO Pro
Smart Meta Box (Gutenberg, Classic Editor & Elementor)
The Smart Meta Box is the single most-used surface in Skill WP SEO Pro and the place where 90% of your daily SEO work happens. It is engineered as an editor-agnostic React panel that auto-detects whether you are writing inside Gutenberg's block editor, the Classic TinyMCE editor, or Elementor's edit screen, and then mounts the appropriate UI shell while sharing the same underlying data store, the same scoring engine, and the same AI helpers. This means an article started in Gutenberg, continued in Classic, and finished in Elementor will carry identical SEO data the entire way, with no synchronization drift, no duplicate meta rows, and no surprise overwrites.
At the top of the meta box sits the Focus Keyword input, which is far more than a passive text field. As you type, the plugin queries its local keyword index and the configured AI provider to surface synonyms, semantic variants, and related entities you should naturally weave into the body. The SEO Title and Meta Description fields render a live, pixel-accurate preview of how the snippet will appear in Google's desktop and mobile SERPs, with color-coded warnings the moment you exceed safe character or pixel widths. A small but powerful detail is that the title preview uses Google's actual rendering font and width algorithm, not just a character count, so you never publish a title that gets truncated to an awkward ellipsis in production.
Below the snippet, the meta box exposes the Canonical URL override, the full Open Graph and Twitter Card editor with auto-generated share images, granular Robots directives (noindex, nofollow, noarchive, nosnippet, max-snippet, max-image-preview, max-video-preview), a per-post Breadcrumb label override, and a Reading Time + Word Count indicator. All of these settings are persisted to post meta with a single, well-namespaced key prefix, so migrations, exports and developer queries stay clean. The entire panel is keyboard-navigable, screen-reader friendly, and respects WordPress's color scheme, which matters when editorial teams have heavy accessibility requirements.
The Smart Meta Box also integrates with the AI Content Optimizer described in the next section, so every change you make in the meta box immediately re-runs the 18 SEO checks and updates the live score ring. There is no save-and-refresh dance — feedback is instantaneous. For multilingual sites, the meta box is fully WPML- and Polylang-aware: it shows you per-language fields, prevents you from accidentally mixing locales, and lets the AI helpers respect the target language when generating titles, meta descriptions, or FAQ blocks. For WooCommerce stores, the meta box adds product-specific fields like brand, GTIN, MPN, and offer availability, which feed directly into the Product schema emitter.
Finally, every action inside the meta box is logged to the audit trail when role-based permissions are enabled, so site owners can see exactly who changed which SEO field on which post and when. Combined with the reversible per-post snapshot system, this means editorial mistakes are recoverable in seconds rather than requiring a database restore. The Smart Meta Box is intentionally calm, fast, and out of the way — but every pixel of it is engineered for the long, repetitive work of producing high-quality SEO at scale.
AI SEO Content Optimizer with 18-Point Live Scoring
The AI SEO Content Optimizer is the headline capability of version 7.2.9 and the feature that, more than any other, separates Skill WP SEO Pro from the previous generation of SEO plugins. It runs a real-time, editor-side scoring engine that evaluates your draft against 18 weighted checks every time you stop typing for more than 400 milliseconds. The result is rendered as an animated circular score ring from 0 to 100, color-graded into four readable buckets: Excellent (85+), Good (65–84), Needs work (40–64), and Poor (below 40). The ring sits in a sticky panel that stays visible as you scroll, so you always know where you stand without breaking flow.
The 18 checks are not a vanity list. Each one is grounded in either a confirmed Google ranking factor, a documented best practice from major SEO studies, or a measurable readability principle. They include focus keyword in the SEO title, focus keyword in the first 100 words, focus keyword in at least one H2 or H3, focus keyword in the URL slug, focus keyword in the meta description, keyword density between 0.5% and 2.5%, title length between 30 and 60 characters, meta description length between 120 and 160 characters, slug length under 75 characters with stop-words stripped, at least one outbound dofollow link to an authority domain, at least two internal links to related content, at least one image per 300 words, 100% image alt-text coverage, exactly one H1, an H2 every 300 words, a Flesch Reading Ease score of 60 or higher, an average sentence length of 20 words or fewer, and paragraph length capped at 150 words.
Each check is clickable. Click a failing check and the editor scrolls to the offending location and highlights it. Click the small ✨ icon next to many checks and the AI provider is invoked to fix the problem in place — generate a better title, rewrite the introduction with the keyword in the first sentence, add an internal link to a related article, or split a long sentence into two. The AI never silently overwrites your work; it always shows a diff preview and asks for confirmation before applying changes, so authorship stays with the author.
What makes the optimizer truly different from older plugins is that the scoring weights are tunable per-site and per-post-type. A news publisher who values short, punchy paragraphs can lower the weight of the paragraph-length check. A long-form B2B SaaS blog can raise the weight of internal linking and reading-ease. The defaults are sensible and based on aggregated ranking-correlation data, but no two sites are identical and the optimizer respects that. The weights are stored in a versioned options table, so changes can be rolled back and so site-wide A/B tests of different scoring profiles are possible.
Persistence matters. Every score, every focus keyword, every meta override, and every AI-applied edit is written to post meta with a clear schema, which means dashboards, reports, WP-CLI commands, and external BI tools can read SEO health across the entire site without scraping HTML. The optimizer also writes a small JSON event to the audit log every time a score crosses a threshold, which is invaluable for editorial managers trying to spot quality drift in long-running content programs.
In production we routinely see editorial teams lift their average post score from the mid-50s to the mid-80s within three to four publishing cycles simply by acting on the optimizer's prompts. That, in turn, correlates with measurable click-through-rate improvements in Google Search Console within four to eight weeks, because better titles, meta descriptions, internal links, and reading ease are not abstract metrics — they are the exact levers that move CTR, dwell time, and ranking.
AI Writer, Editor Assistants & Brand Voice
Skill WP SEO Pro ships with six AI helpers that live directly inside the editor sidebar, each callable with a single click and each governed by your configured Brand Voice. The six are: Generate SEO Title, Generate Meta Description, Generate Article Outline, Generate FAQ Block, Rewrite Introduction, and Rewrite Selected Paragraph. They share a single prompt engineering layer that injects four pieces of context into every call — your brand voice, the post's focus keyword, the target language, and a topical context window built from the existing draft. This shared context is the reason output from one helper composes cleanly with output from the others: a title generated in step one fits the outline generated in step three, which fits the FAQ generated in step five.
The Brand Voice itself is a 1–3 sentence description you write once and store at the site level. Good examples look like: "We write for solo founders and bootstrapped SaaS owners. Tone is direct, calm, and slightly contrarian. No fluff. Examples and concrete numbers always. American English." The AI helpers append this voice to every system prompt, which means generated copy stops sounding like generic ChatGPT and starts sounding like you, your team, your audience. For larger publishers, the brand voice can be overridden per-post-type — a tutorials post type uses one voice, a news post type uses another, a product launch post type uses a third.
Generate SEO Title produces five candidate titles, each scored for keyword inclusion, character length, and emotional pull. You pick one or hit regenerate. Generate Meta Description does the same for the snippet, with a strong bias toward front-loading the focus keyword and ending with a soft call to action. Generate Article Outline produces a full H2/H3 structure organized into semantic clusters, with question-style headings prioritized for AEO and People Also Ask capture. Generate FAQ Block produces 4–8 schema-ready Q&A pairs derived from real PAA queries and inserts them with FAQ schema attached.
Rewrite Introduction is the helper most editors fall in love with. It takes your existing introduction (or generates one from the outline if the intro is empty), and rewrites it into a 3-sentence hook + thesis + keyword structure that reliably scores well on the AI Content Optimizer. Rewrite Selected Paragraph works on whatever text you have highlighted, with tone presets — clearer, shorter, more authoritative, more conversational, more technical, more beginner-friendly. Selection-based rewriting is more precise than "rewrite the whole article" workflows because you keep voice and judgment, and you only delegate the mechanical polish.
Every helper respects the daily token budget configured in AI Settings, so AI spend is predictable. Every call is logged to the audit table with timestamp, post ID, provider, prompt size, completion size, and a per-call cost estimate, so a finance team can reconcile AI cost against editorial output. And every helper has a one-click "undo" because the diff preview shows you exactly what will change before you accept it. The result is a writing experience where AI is genuinely useful — not a black box, not a replacement for the writer, but a tireless second pair of hands that catches the boring, repeatable mistakes and frees the writer to do the work only humans can do.
AI Keyword Generator & Deep Keyword Research
The Keyword Research module is built around a single principle: a seed topic should produce a structured, decision-ready keyword brief in under sixty seconds, not a 5000-row CSV that takes a junior SEO an afternoon to clean. You enter a seed topic — say, "ecommerce returns automation" — and the engine returns a JSON array of 30 high-value, non-duplicate keywords spanning short-tail, long-tail, and natural question-style queries. Each row is a structured object with the keyword itself, the search intent (informational, commercial, transactional, or navigational), an estimated difficulty score from 1 to 100, a low/medium/high volume estimate, a low/medium/high CPC estimate, a type tag (short-tail, long-tail, or question), and a cluster label that groups related keywords into ready-made content silos.
The intent tagging is what makes the output immediately usable. Most legacy keyword tools return a flat list and force you to manually triage each row for intent before you can plan content. Skill WP SEO Pro's research engine front-loads this work because the underlying prompt is explicit about returning at least eight question-style queries that AI engines are likely to cite, and because the AI is instructed to think about each keyword's place in a real buyer journey before tagging it. The result is a brief where transactional keywords cluster naturally into product/pricing pages, commercial keywords cluster into comparison pages, informational keywords cluster into tutorial and guide pages, and navigational keywords are filtered out as not worth your content investment.
The cluster labels are the second piece of magic. When the engine returns 30 keywords grouped under labels like "returns policy templates", "automated refund triggers", "reverse logistics SaaS", "return fraud prevention", and "return rate benchmarks", you have not just keywords — you have the topical map of an entire pillar-and-cluster content silo. One click on a cluster label opens a draft content brief with an AI-generated H2/H3 outline that covers every keyword in the cluster, internal-link suggestions to existing posts that already touch the topic, and a recommended internal anchor strategy.
Context is preserved across runs. If you run keyword research with an existing article ID attached as topical context, the engine uses up to 2000 characters of that article's body to bias the output toward keywords that build on, rather than repeat, what you already publish. This is critical for established content programs where the last thing you want is two posts competing for the same query (keyword cannibalization). The plugin even surfaces cannibalization warnings inline — if a returned keyword already has a ranking page on your site, it is flagged with the existing URL and current rank, so you can decide whether to refresh the old post or write a new one.
Exports are clean. You can export the full research result as JSON, CSV, or a pre-built WordPress draft post with the focus keyword set, an outline pre-filled into the editor, and an FAQ block attached. This last export is the one that bridges keyword research to content production with zero friction. Combined with the AI Writer helpers described above, the typical loop is: seed → 30 keywords in 60 seconds → pick a cluster → generate brief → generate intro → write the body → publish. Sites that run this loop weekly report doubling their indexed pages within a quarter without doubling their writing team.
Finally, the engine is provider-agnostic. DeepSeek, OpenAI, Anthropic, Gemini, Perplexity, and OpenRouter all produce valid output because the prompt is engineered defensively, the response is always validated against a strict JSON schema, and any unparseable AI response is captured with the raw text returned to the UI for human review. There are no silent failures, no half-results, no mysterious "keyword research returned nothing" empty states. Either you get a structured brief, or you get a clear error with the raw provider response so you can debug.
XML, News, Image & Video Sitemaps and JSON-LD Schema
Sitemaps and structured data are the unglamorous plumbing of SEO — when they work nobody thinks about them, and when they break, ranking quietly collapses. Skill WP SEO Pro treats both as first-class, fully-automated subsystems with sensible defaults that work the moment you activate the plugin and deep customization for the edge cases. The XML sitemap engine generates a sitemap index plus per-post-type child sitemaps, automatically split when a single sitemap exceeds 1000 URLs (well under Google's 50,000-URL hard limit but tuned for fast crawling), and includes lastmod, changefreq, and priority hints derived from your actual publishing activity rather than hand-tuned guesses. The sitemap is regenerated on every publish and pings Google and Bing automatically within five seconds.
The Google News sitemap is a separate emitter activated for any post type tagged as news. It includes only the last 48 hours of news content, exactly as Google's News Publisher Center spec requires, with publication name, language, genres, and news keywords. Image sitemaps are emitted with every relevant URL, including image caption, title, geo-location (when GPS EXIF data is present), and license URL — all of which feed Google's image-search ranking signals. When Skill WP SEO Media Pro is active, a video sitemap is wired in automatically with VideoObject metadata for every embedded YouTube, Vimeo, Wistia, or self-hosted video.
On the structured data side, the plugin ships emitters for over 25 Schema.org types: Article, BlogPosting, NewsArticle, Product, Review, AggregateRating, FAQ, HowTo, LocalBusiness, Organization, Person, BreadcrumbList, Event, Course, Job Posting, Recipe, SoftwareApplication, Book, Movie, VideoObject, Service, Offer, Brand, WebSite, and WebPage, with more added each release. The emitter is context-aware: it inspects the current post's type, taxonomy, custom fields, and content blocks to decide which schema types are appropriate, and it composes them into a single JSON-LD graph node with proper @id references, so the resulting schema validates cleanly in Google's Rich Results test.
Every emitter is overridable per-post via the Smart Meta Box's Schema tab. For Product schema on WooCommerce stores, the emitter pulls SKU, GTIN, MPN, brand, price, availability, and aggregate review data from the product's native fields with no extra configuration. For LocalBusiness schema, the global settings page exposes NAP (name, address, phone), opening hours by day-of-week, service area, payment methods accepted, and price range. For Organization and Person schema, the Knowledge Graph editor lets you list sameAs URLs pointing to social profiles, Wikipedia, Wikidata, Crunchbase, and other authority sources, which is the single biggest lever for getting your brand into Google's Knowledge Panel.
Breadcrumbs are handled as both semantic HTML5 (the on-page breadcrumb trail) and JSON-LD BreadcrumbList schema, with one shortcode `[swsp_breadcrumbs]` or one PHP action `do_action('swsp_breadcrumbs')` to render them anywhere in your theme. The HTML output is styling-agnostic and uses semantic ` ` and ` ` elements with proper aria attributes, so themes and accessibility audits both stay happy. Site verification meta tags for Google Search Console, Bing Webmaster, Yandex, Pinterest, and Baidu are injected from a single settings screen with one-click verification, so the chore of pasting verification snippets into header.php is eliminated forever.
The end-state of this subsystem is that a freshly-activated Skill WP SEO Pro on a brand-new WordPress site immediately ships a correct, valid, comprehensive sitemap and structured-data layer with zero configuration. Advanced sites with custom post types, multilingual setups, WooCommerce stores, news verticals, and video libraries all get the same defaults plus the deep override hooks they need. We have measured first-time-correct Rich Results test pass rates above 98% across hundreds of test sites, which is meaningfully higher than the legacy SEO plugin we benchmarked against.
GSC + GA4 Dashboard, Rank Tracker, PageSpeed & Multilingual
The native Google Search Console and GA4 dashboard is the feature that ends the tab-switching tax most SEOs pay every day. You connect via OAuth (no key-copying, no service account JSON files), pick the GSC property and the GA4 property that match the site, and within two minutes the dashboard inside WordPress is populated with the data that actually matters: top organic queries with impressions, clicks, CTR and average position; pages losing rank week-over-week with the magnitude of the loss; new keywords entering the top 10 and top 3; country and device breakdowns; and the highest-value CTR opportunities — pages with high impressions but a CTR materially below the cohort median, which are the easiest wins because the ranking already exists and only the snippet needs improvement.
The dashboard is not just a pretty wrapper around the GSC API. It joins GSC query data with on-site signals like the current SEO title, current meta description, current schema state, and current SEO score from the Content Optimizer, so when you spot a CTR opportunity you are one click away from the meta box that fixes it. The same joined view powers the weekly Editorial Health Report emailed to administrators, which highlights the five most-improved pages, the five most-declined pages, and the five highest-opportunity pages, with a one-click action on each.
The Rank Tracker module tracks keyword positions in Google and Bing across desktop and mobile for any list of keywords you care about. Positions are checked daily from rotating geographic locations, cached, and surfaced as line charts so you can see ranking trends over weeks and months rather than fixating on day-to-day noise. Keywords can be grouped into pillars (often matching the clusters from the Keyword Research engine), so a marketing manager can answer the question "how is our returns automation pillar trending?" with a single chart instead of a spreadsheet.
PageSpeed integration runs Google's PageSpeed Insights and Lighthouse audits on demand, with results cached for 24 hours so you do not hammer the API. Mobile and desktop scores are tracked over time and tied to the URLs they belong to, so a content team can see exactly which post launches caused regressions in Core Web Vitals — invaluable when a heavy hero image or a poorly-written analytics snippet quietly degrades performance.
Multilingual SEO is fully supported through native WPML and Polylang integrations. Hreflang clusters are generated automatically across all language variants of every URL, with the correct x-default fallback and the correct self-referencing tags that prevent hreflang errors in Search Console. Per-language meta titles, meta descriptions, OG tags, and FAQ blocks are stored per-language with no leakage between locales, and the AI helpers respect the active language so a French post never receives English-language meta-description suggestions. Hreflang errors, when they do occur, are reported in the dashboard with a one-click fix that re-syncs the language cluster.
Local Business sites get a dedicated LocalBusiness section with full NAP schema, opening-hours, service-area, and review aggregation. Author and Category archives get dedicated SEO meta and rich Article/CollectionPage schema. Role-based permissions let you give editors meta-box access without giving them settings access, restrict AI key visibility to administrators only, and audit who changed what and when. This last detail matters more than it sounds: in any team of more than three editors, the question "who changed the meta description on this post?" comes up weekly, and the audit log answers it in two clicks.
Skill WP SEO Tools Pro
Redirection Manager with Regex, Geo & AI 404 Suggestions
The Redirection Manager is the most operationally critical module in Skill WP SEO Tools Pro because broken URLs leak ranking equity faster than almost any other technical SEO mistake. The module supports the full HTTP redirect status code spectrum — 301 Moved Permanently, 302 Found, 307 Temporary Redirect, 308 Permanent Redirect, 410 Gone, and 451 Unavailable for Legal Reasons — and exposes each option in plain English so even non-technical content managers can pick the right one without reading the HTTP RFC. The default is 301, which is correct for 95% of real-world cases, but the choice is yours.
Beyond the basic source-to-target mapping, every redirect supports five powerful conditional modifiers that lift the module well above the capability of free redirect plugins. You can scope a redirect to specific user-agents (route mobile-only URLs), to specific countries via Geo-IP lookup (route a US visitor to /us/ and a UK visitor to /uk/), to specific referrers (route inbound traffic from a campaign URL to a tailored landing page), to logged-in versus logged-out users (route members to a member-only resource and visitors to a sales page), or to specific user roles (route admins to a staging variant of a page).
Regex redirects with named capture groups handle the migration cases that destroy weekend after weekend of SEO time when done manually. A pattern like `^/blog/(\d{4})/(\d{2})/(.+)/?$` rewriting to `/articles/$3` collapses an entire date-based URL structure into a slug-based one in a single rule. The regex engine is the standard PCRE2 engine PHP ships with, and the UI includes a live tester that lets you paste a sample URL and see exactly what the rule will do before saving. Regex mistakes are the most common cause of redirect-loop disasters, and the tester is the single biggest reason this module's bug-report rate is meaningfully lower than the legacy alternatives.
AI-powered 404 redirect suggestions are the headline capability. Every 404 logged by the 404 Monitor is automatically fed through the configured AI provider with the site's full URL inventory as context, and the AI proposes the closest matching live URL with a confidence score from 0 to 100. The dashboard renders these suggestions as one-click approve buttons next to each 404, so a site owner can clear a week of accumulated 404s in ten minutes rather than ten hours. The AI is conservative by design: suggestions below a configurable confidence threshold (default 70) are flagged for manual review rather than auto-applied, and you can always reject a suggestion to teach the system that two URLs you might think are similar are actually unrelated.
Bulk import and export use a simple CSV format with columns for source, target, status code, regex flag, and optional condition fields. Migrations from Yoast Premium Redirects, Rank Math Pro Redirects, Redirection by John Godley, and Safe Redirect Manager are supported with one-click format detection and conversion. A per-redirect hit counter and last-hit timestamp let you see at a glance which redirects are actually carrying traffic — usually 10% of your redirects handle 90% of the traffic, and the rest can be pruned during quarterly cleanups without measurable impact.
Performance matters because redirects run on every request that does not match a real URL. The module uses an in-memory cached lookup table refreshed only when redirects change, not a database query per request. On a site with 50,000 redirects the per-request overhead is under one millisecond. The cache invalidates immediately when you save, delete, or import redirects, so there is never a stale-cache window where a freshly-deleted redirect continues to fire. For sites on multi-server hosting, the cache is shared across servers via the object-cache layer, so all nodes see redirect changes within seconds rather than at the next individual cache TTL.
Finally, every redirect change is logged to the audit trail. In a team setting, the question "who deleted the redirect for the old pricing page?" becomes answerable in one click, which is the difference between a five-minute investigation and a five-hour panic when traffic to a redirected URL suddenly disappears.
404 Monitor, Smart 404 Page & Broken Link Checker
The 404 Monitor is the surveillance layer that feeds the Redirection Manager. Every 404 response served by WordPress is captured in a dedicated log table with the requested URL, the referrer (so you can tell whether the broken link came from inside your site or from an external source), the user-agent (so you can distinguish real visitors from crawlers), the visitor's IP (anonymized to /24 for GDPR), and the timestamp. The log is paginated, searchable, sortable, and exportable, with a default retention window of 90 days, configurable from 7 days up to 2 years depending on your storage budget and audit requirements.
The log surfaces three insight panels that turn raw 404 noise into actionable signal. The first panel groups 404s by URL pattern, so 500 individual 404s on different `/old-product/sku-*` URLs collapse into a single "500 hits on /old-product/* pattern" row that you can fix with one regex redirect instead of 500 individual rules. The second panel groups 404s by referrer, so you can quickly see which external sites are linking to broken URLs on your domain and reach out for link reclamation (often the single highest-ROI link-building activity for any established site). The third panel groups 404s by entry-page-to-error-page path, so you can spot internal broken links inside your own navigation, sidebars, or footers — these are the worst 404s because they hurt user trust and waste crawl budget.
Smart 404 is the user-facing complement to the monitor. When toggled on, it replaces the default WordPress 404 page with a custom template that shows three AI-suggested live URLs the visitor most likely meant, plus a prominent search box, plus links to your most popular pages. The AI suggestions are generated on the fly from the failed URL and your sitemap, cached for 30 days to keep response time fast, and refreshed when site content changes. In our measurement across customer sites, Smart 404 recovers between 18% and 42% of visitors who would otherwise have bounced — a meaningful conversion-rate uplift for sites that have inherited broken backlinks from old domain migrations or product renames.
The Broken Link Checker is the proactive companion to the reactive 404 Monitor. It runs a scheduled, full-site crawl of every internal and outbound link in your published content (configurable cadence: daily, weekly, monthly), and reports broken links (HTTP 4xx and 5xx), redirected links (so you can update to the canonical URL and avoid the redirect hop), soft 404s (pages that return 200 OK but display a "not found" message — often the worst kind of broken link because they hide), and slow links (over 5 seconds to first byte, which hurts user experience and signals neglect).
Each finding has a one-click bulk-fix action: replace the broken URL across all posts where it appears, mark the link as nofollow, remove the link entirely, or open the post in the editor for manual review. Replacements are atomic and reversible — every bulk-fix writes a snapshot to the audit log, and the rollback button restores the previous state in seconds if you discover the bulk-fix was too aggressive. For sites with thousands of posts, the bulk-replace is the single most time-saving feature in the entire suite: cleaning up a single decommissioned external link that appears in 200 old posts takes ten seconds instead of an afternoon.
Crawl performance is tuned to respect your server. The default concurrency is 4 parallel requests with a 250ms delay between requests on the same host, which keeps crawl impact well under 1% of typical server load. Outbound HTTP requests use the WordPress HTTP API with a 10-second timeout and respect robots.txt, so the checker behaves like a polite SEO tool and not a denial-of-service vector against the linked sites. All findings are deduplicated by URL across the entire site, so a link broken in 100 posts shows up as one finding with a count of 100, not 100 separate findings.
Bulk Editor, Importer & Permalink Tools
The Bulk Title & Meta Description Editor is the productivity superpower for any editorial team migrating to Skill WP SEO Pro from another SEO plugin, or for any site that has accumulated thousands of posts with inconsistent or empty SEO meta. The editor presents your entire post inventory as a spreadsheet-style grid with editable title and meta description columns, plus filterable columns for category, tag, custom post type, author, publish date, focus keyword, current SEO score, and any custom field you choose to expose. Inline edits save automatically as you tab between cells, with optimistic UI updates and a small revert-this-edit affordance on each row in case you mistype.
The grid scales to 100,000+ posts thanks to virtual scrolling — only the visible rows are rendered in the DOM, and data is fetched in pages of 100 with a 2-second debounce so a wide-scrolling user does not trigger a flood of AJAX requests. Sorting and filtering happen server-side via indexed queries, so a filter like "posts in the Tutorials category with no meta description" returns in well under a second even on very large sites. Column visibility is configurable per-user, so a content editor can hide developer columns and an SEO lead can show all of them.
The headline capability is AI bulk-rewrite. You filter the grid to a subset (for example, all 800 posts in the Reviews category with an empty meta description), click the Bulk Rewrite button, choose a tone preset and a length target, and the AI generates new meta descriptions for every row in the filter using the post's existing content as context. A diff preview shows you what will change before any writes happen, and you can deselect individual rows that you would rather hand-write. The rewrite respects per-row settings — if a row has a manually-set meta description, it is skipped unless you tick "include manually-set rows" in the bulk action dialog.
The Importer module handles migrations from Yoast SEO, Rank Math, and All in One SEO with zero data loss. It reads focus keywords, SEO titles, meta descriptions, OG/Twitter card data, schema overrides, robots directives, redirects, noindex flags, breadcrumb labels, canonical overrides, and per-post sitemap inclusion flags from the source plugin's post meta and option tables, maps them to Skill WP SEO Pro's schema, and writes them atomically inside a transaction so a failure mid-import never leaves your site in a half-migrated state. Before any writes happen, a reversible snapshot is taken — the rollback button restores the previous state in under 30 seconds even on large sites.
Migration order matters. The recommended sequence is: run the importer's dry-run first to see the diff report, fix any flagged anomalies (usually focus keywords that contain newline characters or meta descriptions over 320 chars), run the actual import, validate a 10-post sample by spot-checking the meta box, then disable (do not delete) the old SEO plugin. Leaving the old plugin disabled but installed for two weeks gives you a fallback if any edge-case post turns out to need attention. After two weeks of clean operation, the old plugin can be safely deleted.
Permalink & Slug Tools fix the long tail of URL-quality issues that accumulate on any site that has been running for more than two years. The bulk slug-rewrite operation strips stop words (the, a, of, to, in, on, for, with), normalizes hyphens, enforces lowercase, transliterates non-ASCII characters into clean ASCII slugs (with proper handling of German umlauts, Polish characters, Vietnamese tones, and so on), and automatically creates a 301 redirect from the old slug to the new slug for every change. Trailing-slash normalization is one toggle: either every URL ends in a slash, or no URL does, with consistent 301s either way to prevent the duplicate-content trap of having both `/about` and `/about/` indexable. Every change is logged, every change is reversible, and the bulk operations are gated behind a confirm-and-preview dialog because a wrong slug change is the single fastest way to break a site's organic traffic.
Robots.txt, .htaccess Editor, HTML Sitemap & Schema Builder
The visual Robots.txt editor takes one of the most error-prone files in WordPress administration and turns it into a safe, friendly interface with syntax highlighting, real-time validation against the robots.txt spec, and a one-click library of common templates. Want to block all AI crawlers but allow Google? One click. Want to allow GPTBot but block ClaudeBot? One click. Want to lock down a staging site from all crawlers? One click. The templates ship with detailed inline comments explaining what each directive does and why you might want it, so a content editor learns the spec while making correct choices.
Every save creates a timestamped backup of the previous robots.txt in a versioned options table, with a one-click rollback that reverts to any previous version. The version history shows who saved the change, when, from which IP, and what the diff was — invaluable for the postmortem when somebody accidentally pushes a `Disallow: /` to production and tanks the crawl budget. The editor also previews exactly how Googlebot, Bingbot, and major AI crawlers will interpret the current ruleset against a sample URL, so you can verify your intent before saving.
The .htaccess editor follows the same safety-first design with syntax highlighting for Apache directives, validation against the most common rewrite-rule mistakes (missing flags, unescaped dots, missing trailing slashes), and pre-save dry-run validation that catches over 90% of breaking errors before they reach the live server. Templates cover the common SEO use cases: force HTTPS, force www or non-www, gzip compression, browser caching headers, blocking hotlinking, blocking bad bots, redirecting old domains, and serving WebP/AVIF with proper fallback. As with robots.txt, every save creates a timestamped backup and rollback is one click.
The HTML Sitemap module gives you a human-readable sitemap that lives at a URL you choose (default `/sitemap.html`) and can be embedded in any post or page via the shortcode `[swsp_html_sitemap]`. Unlike XML sitemaps which exist only for search engines, the HTML sitemap is for visitors — it helps users find content that does not surface naturally in your top navigation, and it gives your internal-linking graph a healthy dose of automatic cross-links from a high-authority footer-linked page. The output is grouped by post type and optionally by taxonomy or by month, with configurable maximum items per group and pagination for large sites.
The Schema Builder is the deepest module in Skill WP SEO Tools Pro and the one that gives developers control over JSON-LD without writing code. It is a drag-and-drop builder where you select a Schema.org type from a searchable picker (every Schema.org type is available, not just a curated subset), drag the properties you want to populate into the canvas, and bind each property to a data source — a static value, a post field, a post meta key, an ACF field, a taxonomy term, a user field, or a computed expression like "the URL of the first image in the post content." The builder generates valid JSON-LD that can be emitted globally, per-post-type, per-taxonomy, or per-individual-post based on rules you define.
Validation against Google's Rich Results requirements runs every time you save a schema template, with clear error messages pointing to the specific required property that is missing or the specific property value that violates the spec ("Product.priceCurrency must be a 3-letter ISO currency code"). A live preview shows the generated JSON-LD against a sample post of the matching type, so you can verify the output before the template goes live. Templates can be exported as JSON, version-controlled in your dotfiles or a Git repo, and re-imported into other sites — invaluable for agencies maintaining schema standards across a portfolio of client sites.
The combination of these four modules — robots.txt, .htaccess, HTML sitemap, and Schema Builder — gives a site owner full control over the four most technical aspects of on-site SEO without ever opening an SSH session or writing a line of code. For developers, the same modules expose action hooks and filters at every meaningful extension point, so custom requirements that are not covered by the UI are always reachable through code.
Database Backup, Website Files Backup, Theme Cleanup & User Role Manager
The Database Backup module produces a full, restorable dump of your WordPress database — every table, every row, every index, every stored procedure if present — compressed with gzip and named with a timestamp and a hash of the WordPress version so you always know exactly what you are restoring. The dump uses native mysqldump-equivalent SQL generation written in pure PHP, which means it works on shared hosting where shell access is unavailable and the actual mysqldump binary is not callable. For very large databases, the dump is streamed in chunks so memory usage stays bounded regardless of database size, and the process can be interrupted and resumed without corrupting the output.
The Website Files Backup module captures the full filesystem under wp-content (or any custom path you specify) with the same chunked-streaming approach. The output is a standard zip file readable by any zip tool, with selectable subsets so you can choose to back up only uploads, only plugins, only themes, only must-use plugins, or any combination. Combined backups (database + files) produce a single archive that can be restored into a completely fresh WordPress install with one command, which makes Skill WP SEO Tools Pro a viable disaster-recovery solution for any site that is not already on a managed-WordPress host with built-in backups.
Destinations matter as much as the backup itself, because a backup stored only on the server it backs up is not really a backup. The module ships with native integrations for local storage (default, fine for small sites), FTP/SFTP, Dropbox, Google Drive, Amazon S3, Backblaze B2, and any S3-compatible storage like Wasabi or DigitalOcean Spaces. Credentials are stored encrypted in the database with a per-site key, and all transfers happen over HTTPS with proper certificate verification. Schedule presets cover daily, weekly, and monthly cadences with configurable retention so you can keep, for example, the last 7 daily backups, the last 4 weekly backups, and the last 12 monthly backups, automatically purging older ones to control storage cost.
Restore is a first-class capability, not an afterthought. From the Backups dashboard you can pick any backup in any destination, click Restore, see a dry-run report of what will change, and confirm. The restore runs inside a transaction-equivalent mechanism for the database (full backup of current state taken first, restore attempted, automatic rollback if any error occurs) and uses atomic file operations for the filesystem (extract to staging directory, atomic swap, automatic rollback on partial failure). Every restore is logged to the audit trail with operator, source backup, destination, timestamp, and outcome.
Theme Cleanup is the underrated module that quietly improves site performance and security for sites that have accumulated technical debt over the years. It scans for unused theme files (themes installed but never activated), orphan media attachments (images uploaded to the library but never used in any published content), zombie post revisions (the third through hundredth revision of the same post, which WordPress keeps forever by default and which can balloon to tens of gigabytes on long-running sites), expired transients, spam comments older than your retention window, and trashed posts older than your retention window. Each category is reported with disk-space and database-row impact, and cleanup is one-click with a preview-and-confirm dialog.
The User Role Manager is a visual editor for WordPress capabilities. It surfaces every capability registered on your site (the core ~70, plus any capabilities added by other plugins or your theme), grouped into logical buckets (posts, pages, media, plugins, themes, users, options, comments, custom post types). You can edit existing roles, create new roles, clone existing roles as starting points, and assign any capability to any role with checkboxes. Preset role templates ship for common needs: SEO Manager (full access to all SEO settings, no access to plugins or themes), Content Editor (full access to posts and pages, no access to settings), Multilingual Translator (full access to translation features for assigned languages, read-only access to everything else). All capability changes are logged for audit, and changes are reversible.
Together, this module set positions Skill WP SEO Tools Pro as more than just an SEO toolbox — it is a meaningful operational layer for any WordPress site that takes its administration seriously. The same plugin that handles your redirects also keeps your backups current, your filesystem clean, and your team's permissions tight, with a single coherent UI, a single audit trail, and a single license.
Skill WP SEO Media Pro
WebP & AVIF Converter with Lossless & Background Queue
The WebP and AVIF Converter exists because images are the single biggest cause of slow WordPress sites, and the next-generation image formats — WebP introduced by Google in 2010 and AVIF finalized by the Alliance for Open Media in 2019 — routinely cut image file size by 25% to 80% versus equivalent-quality JPEG and PNG while preserving visual fidelity at a level no human eye can detect. The converter automates the boring part of adopting these formats: it walks your entire WordPress media library, transcodes every image into WebP, AVIF, or both depending on your preference, stores the derivatives alongside the originals (the original is never deleted unless you explicitly opt in), and serves the right format to the right browser via a ` ` element with proper fallback chains.
The conversion engine is built on PHP's native Imagick extension when available (faster, supports more formats, supports AVIF with libheif/libavif) and falls back to GD when Imagick is not present (universal compatibility, supports WebP on PHP 7.0+, does not support AVIF). The plugin's Site Health probe detects exactly what your server supports and surfaces the answer in the UI before you start the bulk conversion, so you never start a job that will fail halfway. For AVIF specifically, the requirements are PHP 8.1+ with Imagick built against a recent enough libheif and libavif — most modern managed-WordPress hosts already ship this combination, and most low-cost shared hosts do not.
Quality control is granular without being overwhelming. A single global quality slider (default 82, which is the empirically-validated sweet spot where files are 60–70% smaller than equivalent JPEG with no visible loss of detail) covers most cases. For images where lossless reproduction is required — logos, screenshots, technical diagrams, brand assets — a per-image lossless toggle is available, and the bulk converter supports a smart heuristic that auto-applies lossless to images detected as logos or screenshots (uniform color regions, few unique colors, high-contrast edges) and lossy to photographs (smooth gradients, many unique colors, soft edges). The lossy/lossless decision is one of the most consequential image-optimization choices, and the auto-heuristic gets it right approximately 95% of the time across our test corpus.
Background queue processing is the feature that makes bulk conversion practical on shared hosting. Naive image processors run in a single long-running HTTP request, which times out on hosts with strict execution-time limits and corrupts the conversion job. The Skill WP SEO Media Pro queue splits the conversion into chunks of 20 images, processes each chunk inside a single safe request window, and uses WordPress's Action Scheduler to chain chunks together transparently. The result is that converting a 50,000-image library takes hours of background time instead of one massive timeout, and you can navigate away from the dashboard, close your laptop, and come back to a finished job.
Frontend serving is fully transparent. The plugin hooks into `wp_calculate_image_srcset` and the post content filter to rewrite every ` ` tag into a ` ` element with AVIF source first, WebP source second, and original JPEG/PNG fallback. Modern browsers pick the best supported format automatically; older browsers fall back to the original without any user-facing degradation. CDN integration is built in — Cloudflare, Bunny, KeyCDN, and Cloudinary URL rewrites are emitted automatically when you connect a CDN, so the WebP and AVIF derivatives are served from the edge with proper MIME headers and proper cache lifetimes.
Original-file retention is the safety net that gives nervous site owners the confidence to enable bulk conversion. The original JPEG or PNG is never deleted by default; the WebP and AVIF derivatives sit alongside it in the same upload directory. If you ever decide to revert — perhaps you discover a workflow that depends on the original format, or you change your mind about quality settings — a one-click revert restores the original references across all post content with a single SQL update inside a transaction. The reversibility of this module is the reason we consider it one of the safest large-scale media operations available in the WordPress ecosystem.
AI Image Alt Text, Title, Caption & Bulk Multilingual Generation
Alt text is the most under-loved attribute in WordPress. It is required by WCAG accessibility guidelines, it is a confirmed Google image-search ranking factor, and it is the only signal screen-reader users have to understand what an image conveys. Yet on a typical 5-year-old WordPress site we routinely measure 60% to 90% of images missing alt text entirely, and most of the remaining alt text is either the original camera filename ("IMG_0237.JPG") or the post title repeated word-for-word. The AI Image Alt Text module fixes this at scale.
The module works by feeding every image through a vision-capable LLM — GPT-4o, Gemini 1.5 Pro or 2.0 Flash, and Claude 3.5 Sonnet or newer are all supported — along with three pieces of context: the title of the post the image appears in, a 500-character snippet of the surrounding paragraph, and the locale of the post. The vision model produces an alt text that describes what is actually in the image, contextualized to where the image appears. The difference between context-free alt text ("a person standing on a stage") and context-aware alt text ("Sundar Pichai presenting Gemini 2.0 at Google I/O 2024") is enormous, and it is the difference between alt text that helps SEO and alt text that is just noise.
Bulk processing is the headline workflow. You open the Bulk Alt Text tool, filter to the subset you want to process (the default is "all images with empty alt text in published posts," which is the right starting point for almost every site), pick the target language (auto-detected from the WordPress site locale, overridable to translate alt text into a different language than the post body if needed), and click Generate. The queue processes images at a rate of approximately 600 to 1500 per hour depending on the AI provider and your account's rate limits, with all activity surfaced in a live progress bar. You can leave the page; the job continues in the background.
Per-image manual override always wins. The module never overwrites an existing non-empty alt text unless you explicitly opt in with the "include manually-set alt text" checkbox in the bulk action dialog. This is the right default because some alt text — product names with SKU codes, brand-specific language, technical diagrams with precise terminology — is better written by humans than guessed by AI, and you do not want a bulk operation to clobber that careful work. When you do opt in to overwriting, a diff preview shows you every change before any writes happen.
The same module also auto-fills the image's Title attribute and Caption when those are empty, using complementary prompts that produce title text optimized for image search (concise, keyword-front-loaded, no punctuation tricks) and caption text optimized for human readers (a complete sentence that adds value beyond what the image already shows). These three attributes — alt, title, caption — work together: alt for accessibility and image SEO, title for hover tooltips and additional crawler signal, caption for the visible text below the image when your theme renders captions. Auto-generating all three from one AI call is more efficient than three separate calls, and the prompts share context so the three outputs stay coherent.
Multilingual sites get explicit first-class support. The plugin reads the post locale via WPML or Polylang and instructs the AI to produce alt text in that locale. For a site with 8 language variants of the same post, the AI alt text is generated 8 times in 8 languages, with proper per-locale storage so French alt text never leaks into the German variant. This is the only practical way we know of to achieve 100% multilingual alt-text coverage without hiring 8 translators.
Rate-limit handling is graceful. If the AI provider returns a rate-limit error mid-job, the queue pauses, waits the recommended backoff period (read from the provider's response headers when available), and resumes automatically. A persistent rate-limit failure surfaces a clear UI notification with a Retry button instead of failing silently. The full job state is checkpointed every 50 images, so a server crash or browser close does not lose progress — when you reopen the dashboard, the job picks up exactly where it stopped. This robustness is what makes bulk AI generation viable as a production workflow rather than a one-off experiment.
Video SEO — Schema, Sitemap, YouTube Facade & Lazy Loading
Video SEO is the most under-served corner of WordPress SEO. Most plugins ignore video entirely; the ones that touch it usually do nothing more than emit a hard-coded VideoObject schema with placeholder data. Skill WP SEO Media Pro treats video as a first-class media type with auto-detection across all major platforms (YouTube, Vimeo, Wistia, JW Player, and self-hosted MP4/WebM), automatic metadata extraction (thumbnail, duration, upload date, embed URL, transcript when available), valid Schema.org VideoObject JSON-LD output, a dedicated video sitemap at `/video-sitemap.xml`, and frontend performance optimizations that genuinely improve Core Web Vitals.
Auto-detection works across Gutenberg blocks, Classic editor shortcodes, Elementor video widgets, and raw embed URLs pasted into post content. The plugin parses post content on save, identifies every video reference, and writes a normalized video record to a dedicated table that the schema emitter and sitemap generator both read from. Re-parsing happens automatically when content changes, and a manual re-scan button is available in case you need to rebuild the index after a bulk content import. The detection is intentionally conservative — it never claims a video exists unless it has high confidence, so the schema and sitemap never include phantom entries that would trigger Google Search Console errors.
Metadata extraction uses the public oEmbed endpoints of each platform when available, falling back to direct API calls for richer data when you have configured API keys. For YouTube videos, the extracted metadata includes title, description, thumbnail URL (in three resolutions: standard, high, maxres), duration in ISO 8601 format, upload date, channel name, channel URL, view count at last refresh, and category. For self-hosted videos, the plugin reads the actual file to extract duration, dimensions, codec, bitrate, and (when present) an embedded poster frame. All metadata is cached for 24 hours by default and refreshed via cron, so the schema stays current without hammering platform APIs.
VideoObject schema emission is built to pass Google's Rich Results test on the first attempt. Every required property (name, description, thumbnailUrl, uploadDate) is populated, every strongly-recommended property (duration, contentUrl, embedUrl) is populated when the data is available, and optional but valuable properties (interactionStatistic for view counts, hasPart for chapters, transcript for accessibility) are included when present. The output is composed into the same JSON-LD graph as the post's other schema (Article, BreadcrumbList, Organization) with proper @id references, which is the correct way to emit multiple schema types per page and is the way Google's documentation recommends.
The video sitemap is generated at `/video-sitemap.xml` with one URL entry per post that contains video, and one ` ` child element per video on that page. Required elements (thumbnail_loc, title, description, content_loc or player_loc) are always present; optional elements (duration, expiration_date, rating, view_count, publication_date, family_friendly, tag, category, gallery_loc, requires_subscription, uploader, live, restriction, platform) are emitted when the underlying data is available. The sitemap is referenced from the main sitemap index automatically, and a dedicated Google Search Console submission button in the Video SEO settings page makes the one-time submission painless.
Performance optimization is where Video SEO crosses from "compliance feature" to "this measurably makes my site faster." The YouTube Facade option replaces the heavy YouTube iframe (which loads 500KB to 1.5MB of YouTube SDK JavaScript before any video plays) with a lightweight static thumbnail and play-button overlay. When a visitor clicks the thumbnail, the real iframe is swapped in and autoplay begins. The user experience is identical; the page weight drops by hundreds of kilobytes; the Largest Contentful Paint improves measurably; and the Total Blocking Time drops because the YouTube SDK no longer blocks the main thread on first render.
Self-hosted videos get the same treatment with native lazy loading via the `loading="lazy"` attribute on ` ` elements, with `preload="none"` so the video file is not downloaded until the visitor interacts. For modern browsers, the `playsinline` attribute is added automatically (required for iOS Safari inline playback), and the poster frame extracted from metadata is set as the `poster` attribute (so the first paint is instant). Captions and transcripts, when present in the metadata, are surfaced both to the schema and to the ` ` element on the video, which improves accessibility and gives Google additional text content to index for the video.
CDN Integration, Lazy Loading & Image SEO Best Practices
CDN integration in Skill WP SEO Media Pro is built around a simple principle: connect once, and every image, WebP, AVIF, and video derivative is served from the edge automatically, with no manual URL rewriting and no theme changes. Native integrations ship for Cloudflare, Bunny.net, KeyCDN, and Cloudinary, with a generic S3-compatible integration that works with any CDN that serves from an S3-compatible origin (Backblaze B2, DigitalOcean Spaces, Wasabi, MinIO, etc.). Cloudflare integration uses an API token scoped to Cache Purge only, so the plugin can purge edge cache when you update an image but cannot make any other changes to your Cloudflare zone.
URL rewriting happens in the same content filter that emits the ` ` element. The original WordPress media URL is replaced with the CDN URL at render time, which means no database migration is needed — the post content in the database still references your origin, and the rewrite is a transparent layer on top. This is significantly safer than approaches that rewrite content directly in the database, because if you ever change CDN providers or disable the CDN, no data migration is needed; you simply update the CDN settings and the rewrites change automatically. For sites that already have CDN URLs baked into post content from a previous setup, a one-click migration tool normalizes everything back to origin URLs and the CDN rewrite handles serving.
Cache invalidation is handled correctly. When you update an image, replace it via the Media Library, or regenerate WebP/AVIF derivatives, the corresponding edge cache entries are purged automatically via the CDN's purge-by-URL API. For Cloudflare, individual file purges are used (which are fast and free); for CDNs that charge per purge, the plugin batches purges into single API calls to keep cost down. Bulk operations that affect many images (a sitewide WebP re-conversion, for example) trigger a tag-based purge or a wildcard purge when the CDN supports it, which is much faster than purging thousands of individual URLs.
Lazy loading is enabled by default on all images and iframes, using the native `loading="lazy"` attribute that every modern browser supports without any JavaScript dependency. The plugin adds the attribute automatically to images that appear below the fold (more than one viewport height down from the page top) and intentionally does NOT add it to above-the-fold images, because lazy-loading above-the-fold content actively hurts Largest Contentful Paint. The fold-detection heuristic considers the image's position in the post content, with a configurable threshold for sites where the typical above-the-fold content extends further than the default 800-pixel mark.
For maximum performance, the plugin also emits `decoding="async"` on every image and `fetchpriority="high"` on the single most important image (typically the hero image of the post, identified by being the first content image in the post body). The combination of native lazy loading, async decoding, and selective high-priority fetching for the LCP candidate is the most impactful image-performance change you can make in WordPress, and it routinely lifts Largest Contentful Paint scores from "needs improvement" into "good" without any code changes.
Image SEO best practices baked into the plugin include responsive `srcset` generation at every WordPress-registered image size (so browsers download the smallest image that fits the viewport), proper `sizes` attributes that match the actual rendered width at every breakpoint (so the browser's image-selection logic picks the right `srcset` candidate), automatic conversion of image filenames to SEO-friendly slugs on upload (with stop-word removal and dash normalization), automatic addition of width and height attributes to every image (which prevents Cumulative Layout Shift when images load), and automatic stripping of EXIF data from uploaded images (which protects user privacy by removing GPS coordinates and camera-serial-number data that most users do not realize is embedded in their photos).
The cumulative effect of these defaults on a fresh site is that out of the box, with no configuration beyond activating the plugin, image-related Core Web Vitals scores improve, image-related accessibility scores improve, image-related SEO signals strengthen, and bandwidth costs measurably drop. We have measured 35% to 60% reductions in image-related bandwidth on customer sites after activation, with corresponding improvements in page-load times that translate directly into better engagement metrics and better organic rankings.
Skill WP Optimize Pro
Page Cache — Disk, Per-Device, Per-Role & Preload
Page cache is the single highest-leverage performance feature in any WordPress stack. A request that hits a generated HTML cache file returns in under 50 milliseconds; the same request hitting PHP and MySQL routinely takes 800 milliseconds to 3 seconds on cheap hosting. Skill WP Optimize Pro's page cache is a full-page disk cache with intelligent invalidation, per-device variants, per-role variants, sitemap-driven preloading, and edge-cache header emission for Cloudflare, Bunny, and Fastly. The cache is engineered to be aggressive (cache everything that can be cached) while being safe (never cache content that varies by visitor without proper variant keys).
The cache storage is a tree of plain HTML files under `wp-content/cache/swsp/` organized by URL path, which means a cache hit involves zero PHP work beyond the bootstrap that checks the cache file exists. The bootstrap is a tiny advanced-cache.php drop-in that runs before WordPress loads, so on a cache hit the request never even starts the full WordPress stack. This is the same technique used by WP Super Cache and W3 Total Cache and is the only way to achieve sub-50ms cache hit times on real PHP-based hosting.
Per-device caching produces three separate cache variants — desktop, tablet, mobile — keyed off the User-Agent header. This is the correct approach for any site running a responsive theme that serves different markup or different image sources to different device categories, and it is essential for sites running AMP variants. The device detection uses a fast User-Agent parser optimized for high-throughput cache lookup; the parser handles the common cases natively and falls back to a regex library only for unusual agents. Cache hit rate on the per-device cache is typically 85% to 95% on a busy site.
Per-role caching is the feature that finally lets you cache content for logged-in users without the security disaster of serving one user's account page to another user. The cache is keyed by user role (administrator, editor, author, contributor, subscriber, customer, plus any custom roles), so a logged-in Editor sees an Editor-specific cache and a logged-in Subscriber sees a Subscriber-specific cache, with proper isolation between them. Per-user caching (where every individual user gets their own cache) is also supported but is rarely the right choice because cache hit rates collapse; per-role is almost always what you actually want.
Cache preloading is the workflow that warms the cache after a cold start, after a sitewide purge, or on a regular schedule. The preloader walks your XML sitemap, makes a polite background request to every URL, and ensures every URL has a fresh cache entry before any real visitor arrives. The default preload concurrency is 2 parallel requests with a 500ms delay, which keeps server load minimal even on shared hosting. For sites that publish new content frequently, the preloader is scheduled to run after every publish or update event, so freshly-published content is already cached when the first visitor arrives from a tweet or a Google update.
Edge cache integration uses standard HTTP cache headers (Cache-Control, Surrogate-Control, CDN-Cache-Control) plus CDN-specific purge APIs to coordinate with Cloudflare, Bunny, and Fastly. When you enable Cloudflare edge cache, the plugin emits the proper Cache-Control headers, registers the appropriate cache rules in your Cloudflare zone via the API, and purges the right URLs from the edge when content changes. The end-state is a three-layer cache pyramid — browser cache, CDN edge cache, server disk cache — where 90%+ of requests are served from the browser or the edge and never reach your origin at all.
Cache invalidation rules are the difference between a cache that works and a cache that quietly serves stale content. The plugin's defaults are: post publish or update invalidates the post URL plus the home page plus the relevant taxonomy archives plus the XML sitemap; theme change invalidates everything; comment posted invalidates the commented post's page; WooCommerce stock change invalidates the product page plus the shop archive plus relevant category archives; menu change invalidates everything (because menus appear on every page). Each rule is configurable for sites with unusual requirements, but the defaults are correct for 95% of WordPress sites and require zero tuning.
Critical CSS, Minification, Combination & Unused CSS Removal
Critical CSS is the optimization that takes a render-blocking stylesheet and turns it into a non-blocking one, with measurable improvements to First Contentful Paint and Largest Contentful Paint. The technique is well-understood: extract the minimum CSS rules needed to render the above-the-fold content, inline those rules in a `
Database Optimizer, Heartbeat Control & Cron Optimization
Reclaim bloat from post revisions, transients, and orphaned metadata on a schedule, throttle the WordPress Heartbeat API to cut admin-ajax load, and replace WP-Cron with a real server cron for predictable, low-overhead background processing.
Core Web Vitals Real-User Monitoring & CWV Recommendations Engine
Collect field LCP, INP and CLS from real visitors, segment by device and template, and surface a prioritised, plain-English recommendations queue that tells you exactly which element, script, or image is costing you the most.
Skill WP LLMs Gen Pro
Auto-Generated llms.txt with Granular Post-Type Control
Publish a standards-compliant llms.txt that tells AI crawlers exactly which pages, post types, and docs to read, with per-post-type include/exclude rules so you expose the right content to large language models and nothing else.
llms-full.txt — Full Markdown Content Export & Per-Post Control
Generate a complete, clean Markdown export of your content as llms-full.txt so answer engines can ingest the full article body — with per-post toggles to opt individual pieces in or out of AI training and retrieval.
Skill GEO/AEO AI Pro
AI Overview, SGE & Answer Engine Citation Tracker
Track when and where your brand is cited inside Google AI Overviews, ChatGPT, Perplexity, Gemini and Copilot answers, monitor citation share over time, and discover the exact queries that trigger an AI mention of your site.
Entity Graph Builder, Geo Variants & Brand Voice Trainer
Build a structured entity graph that connects your brand, products, people and places, generate geo-targeted content variants for local relevance, and train a reusable Brand Voice profile so every AI output sounds unmistakably like you.
SKILLSEOLIC
License Key Generation, Domain Locking & Multi-Domain Plans
Issue and revoke license keys from one central panel, lock activations to specific domains, and manage single-site, multi-site and agency plans with clear seat counts and instant deactivation when a project ends.
Enhanced Dashboard (v7.2.9), Plugin Status Grid & REST API Reference
A redesigned 7.2.9 control center shows the live status of all seven plugins in one grid, surfaces update and license health at a glance, and ships a complete REST API reference for headless and automated workflows.
One License. Every SEO Tool You Need.
AI Content Optimizer
18 weighted, clickable checks score every draft in real time with a live ring and one-click AI fixes.
6 AI Writing Assistants
Generate titles, meta, outlines, FAQ blocks and rewrites — all governed by your Brand Voice.
Deep Keyword Research
AI keyword generator with volume, difficulty, trend and SERP analysis surfaces real opportunities.
Schema & Sitemaps
XML, News, Image & Video sitemaps plus rich JSON-LD schema for every content type.
GSC + GA4 Dashboard
Search Console and Analytics, rank tracking and PageSpeed in one in-dashboard view.
Smart Redirects
301/302 with regex, geo rules and AI 404 suggestions, plus a built-in broken-link checker.
WebP / AVIF + AI Alt
Lossless conversion in a background queue and AI-written alt text, titles and captions.
Speed & Core Web Vitals
Page cache, critical CSS, unused-CSS removal and real-user CWV monitoring.
llms.txt & AEO
Auto-generated llms.txt and answer-engine optimization so AI engines cite your site.
Replaces a Stack of Single-Purpose Plugins
Simple Pricing. Every Tool Included.
- ✓1 website
- ✓Core SEO Pro + Tools Pro
- ✓18-point AI optimizer
- ✓XML & schema sitemaps
- ✓Email support
- ✓5 websites
- ✓All 7 plugins included
- ✓AI writer + Brand Voice
- ✓Speed + Core Web Vitals
- ✓llms.txt & AEO
- ✓Priority support
- ✓Unlimited / 100 sites
- ✓All 7 plugins included
- ✓SKILLSEOLIC license manager
- ✓Multi-domain & white-label
- ✓REST API access
- ✓Dedicated support
Trusted by 120,000+ WordPress Sites
“Our average post SEO score went from the mid-50s to the mid-80s in three publishing cycles. CTR in Search Console followed within weeks.”
“Replacing four plugins with one suite cut our page weight and our maintenance time. PageSpeed jumped to 99 on mobile.”
“The llms.txt and AEO tools are a first — we are now getting cited in AI overviews and ChatGPT answers for our key topics.”
Frequently Asked Questions
Yes. The Pro and Agency plans include all seven plugins — Skill WP SEO Pro, SEO Tools Pro, SEO Media Pro, Optimize Pro, LLMs Gen Pro, GEO/AEO AI Pro and SKILLSEOLIC — under a single license and a single dashboard.
Skill SEO Pro Suite is designed to replace your existing SEO and speed plugins. It includes a guided importer for settings and redirects, so you can migrate without losing your data or rankings.
The 18-point optimizer, AI writer, keyword research and AI alt text are built in. You can connect your own AI provider key for unlimited generation, giving you full control over usage and cost.
llms.txt is an emerging standard that tells AI engines which content to read on your site. The LLMs Gen Pro plugin auto-generates both llms.txt and a full Markdown export so answer engines can find and cite your pages accurately.
Yes. The meta box adds product schema fields for WooCommerce, and the whole suite is WPML- and Polylang-aware with per-language SEO fields and AI generation in the target language.
Every plan comes with a 14-day money-back guarantee. If the suite is not right for your site, contact support within 14 days for a full refund — no questions asked.
