SiteWideMCP

Active users, new users, sessions, engagement time, sessions by channel group, and top pages — with week-over-week deltas and Google Ads spend folded in on request.

Who is on the site right now, broken down by page, country, city, device, or platform.

Google Ads spend, clicks, impressions and CPC by campaign, ad group, keyword, or search query — with a day/week/month time series.

Clicks, impressions, CTR and average position, plus top queries and pages. Filter by query, page, country, device — across web, image, video, news, and Discover.

Is this URL indexed? Coverage state, robots and fetch state, last crawl, Google’s chosen canonical vs yours, mobile usability, rich results.

Submitted sitemaps, when Google last downloaded each, warnings and errors, and submitted-vs-indexed counts.

Top Bing queries with clicks, impressions and position. Copilot and part of ChatGPT’s web results run on Bing’s index, so this is visibility Search Console cannot show you.

Daily clicks, impressions, CTR and indexed-page counts from Bing Webmaster Tools.

How much of your site Bing actually holds, plus crawl errors, robots blocks, 404s and 5xxs. The tool that explains why you’re missing from Copilot answers.

Backlink-profile strength (0–100) for any public domain — yours or a competitor’s. Works out of the box, no Ahrefs account needed.

Crawl issues from your Ahrefs Site Audit project: health score, counts by severity, worst issues first. Needs your own Ahrefs key.

Keyword-level rankings from Rank Tracker: position, week-over-week change, volume, traffic, SERP features. Needs your own Ahrefs key.

The full SERP for one tracked keyword — every competitor’s URL with its own DR, UR and traffic. Needs your own Ahrefs key.

Competitive benchmarking against your tracked keyword set: stats, keyword overlap, top pages, top domains. Needs your own Ahrefs key.

Volume, difficulty, CPC, traffic potential and search intent for any list of keywords — no project needed, so you can research a market you don’t rank in yet.

Keyword ideas from one seed: matching terms, related terms, or the questions people actually search.

Domain-level organic traffic, keyword count, backlinks and referring domains for any site — yours or a competitor’s.

Every keyword a domain already ranks for, with position, volume, traffic and intent — unfiltered by anyone’s Rank Tracker setup.

Who actually links to a domain, with each linking site’s own strength — the report behind “how did they outrank us”.

Set up Rank Tracker yourself: create a project for your site. Idempotent — running it twice does nothing the second time.

Add keywords to a Rank Tracker project. Additive only — nothing here can delete your Ahrefs data.

Account totals over a range: spend, impressions, clicks, CTR, average CPC, conversions, conversion value, cost per conversion and ROAS. Reads the Ads API directly, so it isn’t limited to what GA4 imported.

Campaigns with budget, status, bid strategy and performance, ranked by spend.

What people actually typed, versus the keywords you bid on. Filter to terms that spent money and converted zero times — your negative-keyword list, written for you.

Keyword performance with quality score and all three of its components, so you can see whether the problem is the ad, the landing page, or the expected CTR.

The start-here check: crawls your site and returns findings, worst first — each with a severity, the number of pages affected, examples and a fix — plus a preview of how Google would truncate every title and description. No 0–100 score, on purpose.

The raw table behind those findings, a row per page: status and redirect chain, title and its length, description, canonical, H1s, robots directives, hreflang, word count, schema types, link counts, security headers and image attributes.

Every rule in your robots.txt, and specifically whether Googlebot, Bingbot, GPTBot, ClaudeBot or PerplexityBot are blocked. Flags the expensive mistakes — a blanket disallow, blocked CSS and JS, a missing sitemap line.

Your XML sitemaps, followed one index level deep: URL counts, how many lastmods are over a year stale, malformed entries, URLs pointing off-host, and whether any file breaks the 50,000-URL or 50 MB limits.

Fetches your /llms.txt and validates it against the llmstxt.org spec line by line — the file AI crawlers read to find out what your site is for.

The preview step. Reads what is on the page now and hands back every change it would make as a before-and-after with an id — and changes nothing until you say so.

Applies the changes you just reviewed, by id. It cannot write a value that was not shown to you first, and it records the old value on the way past.

Puts it back. Call it with an id to undo that change, or with nothing to see every change made so far — the undo log is browsable, not something you had to keep a receipt for.

Which of your posts and pages have a missing, truncated or duplicated SEO title and description — the worklist the fixes come from.

Everything currently set on one page — title, description, canonical, focus keyword, social fields, schema type — and which of them are writable on your setup.

Is WordPress actually reachable, and if not, which plugin or host setting is in the way? Names the specific blocker — a security plugin, a stripped auth header, a cache serving you the old value back.

Mobile versus desktop Core Web Vitals for one URL, already diffed: both form factors, real-user field data and lab data kept separate, and a plain statement of which device is worse and by how much.

Six months of weekly Core Web Vitals from real Chrome users, with a trend verdict — and up to five sites per call, so you can put yourself next to four competitors. This is the data Google ranks on.

Lighthouse scores and lab Core Web Vitals for any public URL, including your competitors’.

The dates a page actually changed, from the Internet Archive — a change log, not a crawl log. Find the week a competitor repositioned.

What that page said on that date: title, meta description, canonical, H1s and a text sample.

Connects one of your own domains to your token by proving you control it, rather than by typing the name in and being believed.

What the connector can do and what is already wired up for your token specifically — every tool grouped by source, what each one still needs from you, and the questions worth asking first.

The sites connected to your token, with their property and account IDs.

Confirms each of your sites is wired up correctly and names the exact grant that is missing if not.