Ranklogs MCP · Model Context Protocol

Ranklogs MCP: your SEO data, inside the AI you already work in

Connect Ranklogs to Claude or ChatGPT and the AI stops guessing — it reads your live rankings, audits, backlinks and AI-visibility data, then hands you the finished deliverable, not just a chart.

Connect in 2 minutes See what it can do

Works in ClaudeChatGPTCursorany MCP client

Your AI assistant · connected to Ranklogs

get_ranking_alerts get_audit_issues get_new_lost_backlinks

Two of your tracked keywords declined this week, and both trace back to the same audit finding: 27 pages are missing meta descriptions, including the two ranking pages. No backlinks were lost. Fix the metas first — highest estimated click impact — then the slow LCP on /pricing. → Want me to write all 27 meta descriptions now?

Sample conversation · data from the Aurora Coffee Roasters demo project

27 tools · 4 can act 8 data areas ~2 min to connect 0 extra to buy — actions use your plan's limits

What it is

What is Ranklogs MCP?

The direct answer

Ranklogs MCP is a remote Model Context Protocol server that connects your Ranklogs account to AI assistants like Claude and ChatGPT. It exposes 27 SEO tools — and four of them act: add keywords, track prompts, generate content briefs, and run AI-answer gap analyses, all inside your plan's limits.

MCP in 60 seconds: the open standard that lets AI assistants securely use external tools. Like a USB port for AI — plug Ranklogs in once, and your assistant pulls real data into any conversation.

Not just read-only

Ask, then act

The flagship · four tools that act

Spot the gap, then close it — without leaving the chat

27 tools total — 23 read, and four that act: add keywords, track prompts, generate content briefs, and run AI-answer gap analyses. Here's the flow on the Aurora Coffee Roasters demo project.

"Which prompts mention my competitors but not Aurora Coffee Roasters?" Three of your tracked prompts cite competitors and never mention Aurora Coffee Roasters: "best coffee subscription boxes," "specialty coffee roasters near me," and "ethically sourced coffee brands worth trying." "Win the first one." Claude runs the gap analysis (uses 1 of your plan's content analyses, takes about a minute) and returns the brief: the answer to give, what the cited pages have that you don't, the schema to add.

It refuses politely at your caps — the same limits as the app, nothing extra to buy.

The difference

Dashboard = look at the data.
MCP = think with the data.

Ranklogs dashboard

You interpret the numbers

Great for monitoring and deep dives — you do the correlating.

Ranklogs MCP

The AI hands you the conclusion

Built for root-cause, synthesis and production — you leave with the asset.

What you see

Charts and scores → conclusions in plain language

What you do

Interpret it yourself → ask why, the AI correlates every tool

What you leave with

Numbers → the brief, the emails, the fixes

Use cases

What you can do with Ranklogs inside Claude & ChatGPT

Every use case is the same shape: one plain-language prompt, several Ranklogs tools working together, one finished output.

The flagship · AI search visibility

Ask the AI why it never cites you — using your own data, inside the AI itself

Ranklogs tracks which competitors get cited by AI engines for your topics while you're invisible. With MCP, that citation-gap data sits inside the very AI you're trying to get cited by — so you can ask it directly what would earn the citation.

›"Using Ranklogs, why doesn't ChatGPT ever cite my site, and what should I publish to fix that?"

Tools working: get_ai_visibility · get_ai_citations · get_share_of_voice → a publish plan targeting your citation gaps

Root-cause a drop, don't stare at it

The AI correlates ranking alerts, audit issues and backlink changes into one plain-English diagnosis with a fix order.

›"Using Ranklogs, why did my rankings drop this month?"

get_ranking_alerts · get_audit_issues · get_new_lost_backlinks

✎

From finding to finished asset

The audit found 27 pages missing metas; the AI writes all 27, page by page. Same pattern for outreach emails or paste-ready JSON-LD.

›"Using Ranklogs, write meta descriptions for every page missing one."

get_audit_issues · get_outreach_recommendations · generate_schema

⚔

Competitive strategy, reasoned

Not the raw overlap table — the AI reasons about which competitor keywords are winnable and sequences them into a plan.

›"Using Ranklogs, build a content brief from the keyword gap against my competitor."

analyze_competitor_domain · get_keyword_battleground

☰

One prompt replaces the Monday review

Overview, audit, alerts and backlinks synthesized into one prioritized action list with reasoning — not five dashboard visits.

›"Using Ranklogs, what should I prioritize this week?"

get_project_overview · get_audit_summary · get_ranking_alerts

✉

Client reports in minutes

Real account numbers, the branded PDF, and a plain-English executive summary written around it.

›"Using Ranklogs, pull this month's numbers and write the client email."

get_project_overview · generate_report

⛓

Chain it — that's the point

Not 27 separate lookups. The AI chains tools in one conversation: diagnose, verify against the audit, then produce the fix.

›"…now check those pages in the audit and fix what's blocking them."

any tools, combined by the AI as needed

How it works

Connected in four steps

Connect once

Add mcp.ranklogs.com/mcp in your AI app and sign in with your Ranklogs account.

Ask in plain language

"Using Ranklogs, …" — no query syntax, no exports, no pasting CSVs.

The AI calls your tools

It reads your live data — only yours. Usage counts against your plan's normal limits.

You get the deliverable

A diagnosis, a brief, drafted emails, paste-ready schema — ready to use.

27 tools · 8 areas

Every Ranklogs surface, exposed as a tool

23 are read-only — the AI can read and reason over everything. Four act, shown in orange below — and every one of them stays inside your plan's limits.

◉Account & projects

get_account_statuslist_projectsget_project_overview

↗Rank tracking

get_keyword_rankingsget_ranking_alertsget_share_of_voiceget_keyword_battlegroundadd_keywords_to_project

⌕Keyword research

research_keywordsget_topic_clusters

✚Site audit

get_audit_summaryget_audit_issues

⛓Backlinks

get_backlink_profileget_new_lost_backlinksget_outreach_recommendations

✦Competitors & AI visibility

analyze_competitor_domainget_ai_visibilityget_ai_citationsadd_ai_prompt

▤Deliverables

generate_reportgenerate_schema

✎Content optimization

get_content_action_planget_content_targetsget_content_briefget_prompt_gap_briefgenerate_content_briefwin_prompt

Backlink authority is reported as Domain Score (DS) — Ranklogs' own 0–100 authority metric. Tools in orange act inside your plan's limits — everything else only reads.

Setup

How to connect Ranklogs to your AI

Included with every Ranklogs account — usage counts against your plan's normal limits.

Open Settings → Connectors → Add custom connector (claude.ai, Claude Desktop or Claude Code).
Enter https://mcp.ranklogs.com/mcp as the URL.
Sign in with your Ranklogs email and password, then approve the connection.
Ask: "Using Ranklogs, what should I prioritize this week?"

Plain answer first: ChatGPT's standard connectors will not see Ranklogs tools — they only call generic search/fetch. You need Developer Mode.

Open Settings → Connectors → Advanced → enable Developer Mode.
Add a connector with the URL https://mcp.ranklogs.com/mcp.
Sign in with your Ranklogs account when prompted, and approve.

Any MCP-compatible client can connect to https://mcp.ranklogs.com/mcp.

Add the URL as a remote MCP server in your client's config.
If the client doesn't support OAuth sign-in, authenticate with your Ranklogs API key as a Bearer token (find it in your account settings).

Security

Scoped to your account, capped like the app

🔒

OAuth 2.1 sign-in

You log in with your own Ranklogs account. The AI client never sees your password, and you can revoke the connection at any time.

👤

Your data only

The server only ever reads or acts on the connected account's data — never another customer's, never cross-account.

✋

It only acts where you allow it

23 of the 27 tools only read. The four that act — adding keywords, tracking a prompt, generating a content brief, running an AI-answer gap analysis — count against your plan's normal limits and refuse politely once you hit a cap. Nothing happens outside those four.

Ranklogs MCP · FAQ

Questions about Ranklogs MCP

What is an MCP server?

MCP (Model Context Protocol) is the open standard that lets AI assistants like Claude and ChatGPT securely use external tools and data. An MCP server is the plug-in point: connect it once, and the AI can call its tools inside any conversation. Ranklogs MCP is the server that exposes your Ranklogs SEO data this way.

Which AI apps work with Ranklogs MCP?

Claude (claude.ai custom connectors, Claude Desktop and Claude Code), ChatGPT via Developer Mode, and any MCP-compatible client such as Cursor. The connector URL is https://mcp.ranklogs.com/mcp for all of them.

Does it work with ChatGPT?

Yes — via ChatGPT's Developer Mode (Settings → Connectors → Advanced). ChatGPT's standard connector flow only calls generic search and fetch, so it will not see Ranklogs' named tools; Developer Mode is required.

Is Ranklogs MCP included in my plan?

Yes. Ranklogs MCP is included with every Ranklogs account, including the free trial. There is no separate add-on: tool usage simply counts against your plan's normal limits, the same as using the dashboard.

Can the AI change my data or spend my budget?

Mostly no, with four exceptions. 23 of the 27 tools are read-only. Four can act — add keywords to a project, add an AI prompt to track, generate a content brief, and run an AI-answer gap analysis — and each one counts against your plan's normal limits, the same as using the dashboard, and refuses politely once you hit a cap. Nothing else is written: the briefs, emails and fixes the AI produces in conversation stay in the conversation unless you paste them in yourself.

Is my SEO data safe?

You authenticate with your own Ranklogs account over OAuth 2.1, the server only reads that account's data, and you can revoke access at any time. API-key access is available for clients that don't support OAuth.

How many tools does Ranklogs MCP have?

27 tools across eight areas: account and projects, rank tracking, keyword research, site audit, backlinks, competitor research and AI-search visibility, content optimization, plus report and schema generation. 23 are read-only; the other four — adding keywords, tracking an AI prompt, generating a content brief, and running an AI-answer gap analysis — act, inside your plan's limits.

Put your SEO data where the thinking happens.

Every Ranklogs plan includes the MCP server. Connect it to Claude or ChatGPT in about 2 minutes.

Start free trial See pricing