JetOctopus MCP Your tech SEO insights,
one prompt away

Connect JetOctopus to Claude, ChatGPT,
or Cursor. Query your crawl, logs, and Search Console in plain language — then run crawls without leaving the chat.

JetOctopus MCP connecting crawl, logs and GSC to an AI assistant
Without MCP vs with JetOctopus MCP comparison diagram

Your site's data.
Not training data

Most AI assistants answer SEO questions from training data or files you paste in.

JetOctopus MCP is different. It connects directly to your live crawl, your server logs and your Google Search Console. Every answer comes from what's actually happening on your site, right now.

Three sources. One prompt. The cross-source join that used to take a spreadsheet and thirty minutes now takes a query.

What you can ask

Prompt:

"Show pages with GSC impressions but zero Googlebot visits in the last 30 days."

Crawl + logs + GSC joined at the URL level. Pages with search demand that Google is skipping — your highest-ROI fix list.

Prompt:

"Which top-traffic pages does Googlebot crawl but GPTBot and ClaudeBot never touch?"

Pages that Google crawls but GPTBot and ClaudeBot never touch — your blind spots in AI search.

Prompt:

"Start a crawl of my website. Schedule it weekly every Monday at 06:00."

The AI assistant launches the crawl with your last settings, confirms quota cost first and sets the schedule.

Prompt:

"Run a full SEO issue scan and show critical problems first."

Around 100 pre-built checks across crawl, logs, and Search Console, grouped by severity so you know what to fix first.

Prompt:

"Compare my last two crawls and show what changed, starting with what got worse."

Pages that dropped, broke, or got deindexed between two crawls, sorted worst-first so you know where to start.

Read your data. Run your crawls. All from one chat

Analyze — Read

  • Query crawl data, logs and GSC in plain language
  • Run cross-source joins: crawl + logs + GSC at the URL level
  • Scope any query to a saved Segment
  • Run SEO issue scans across 100+ checks
  • Check load time distribution, depth, indexability, internal links

Manage — Write

  • Start, pause, resume, cancel and complete crawls
  • Schedule recurring crawls in any time zone
  • Adjust thread count and page limits on a running crawl
  • Rename and archive crawls and projects
  • Check live quota before starting a run

Before any write action spends quota or stops work, you get an impact preview:
what will run, which project, and how much quota it costs. Nothing happens until you confirm.

Reversible actions like pause, resume, and archive run right away. On shared projects, write actions need Manage access, so view-only teammates keep their full read access.

Combine JetOctopus with your other tools

MCP is an open standard. Your AI assistant can connect to multiple servers at once — JetOctopus alongside Ahrefs, Semrush, SE Ranking or any other MCP-compatible tool.

That means questions like:

"Which pages from my JetOctopus crawl budget leak report have the highest keyword difficulty in Ahrefs?"

"Take the pages Googlebot ignores and show me their search volume from Semrush."

"Which indexable pages with GSC impressions have zero backlinks?"

and many more.

Ahrefs Semrush SE Ranking Google Search Console Google Analytics

Any MCP-compatible tool works alongside JetOctopus in the same chat.

How to connect

Browser sign-in

For Claude Desktop, ChatGPT, Cursor and most desktop clients. Authenticate once via browser. Two clicks, no config files.

Browser sign-in steps: Open Connectors, Add custom connector, Authorize in browser

API token

For CI pipelines, scripts and automated agents. Generate a long-lived token in
your JetOctopus profile. Each token is individually revocable.

Server URL:

https://my.jetoctopus.com/mcp
Generate token and active tokens list

Supported clients:

ChatGPT Claude Gemini Cursor n8n

ChatGPT · Claude Desktop · Claude Code · Cursor · Codex CLI · Gemini CLI · n8n

Any app built on the Anthropic or OpenAI APIs

Step-by-step setup

Privacy

Conversations stay inside your AI client

JetOctopus doesn't see them, log them, or train on them. Only the specific data your assistant
requests on your behalf is transmitted. Access is tied to your account, scoped to your permissions and revocable anytime from your profile settings.

Frequently Asked Questions

What is MCP?
The Model Context Protocol (MCP) is an open standard for giving AI assistants access to external data and tools. JetOctopus implements it so your assistant can query your crawl, logs and GSC directly — instead of working from training data or files you
paste in manually.
Do I need to know how to code?
No. You ask in plain language. The assistant picks the right tools automatically.
Which AI assistants
are supported?
Any MCP-compatible client: Claude Desktop, ChatGPT, Cursor, Claude Code, Codex CLI, Cline, Continue,
Zed, Gemini CLI, n8n and custom apps built on the Anthropic or OpenAI APIs.
Is it included in my plan?
Yes, on all JetOctopus plans. Your plan determines crawl quota and data volume.
What does "read-write" mean
in practice?
Read tools answer questions about your data. Write tools act on it: start a crawl, schedule it, pause it mid-run, change thread count. Anything destructive or quota-consuming shows a preview and asks for confirmation first.
Does the MCP provide access to raw server logs or only
to aggregated reports
and analytics?
No access to raw log files. The MCP only reads the log columns JetOctopus already stores
in our database — the same processed data you see in the app.
Can access be restricted
to specific JetOctopus projects, domains or datasets?
Not available today. If this is something you need — limiting shared-access users to certain projects or actions,
for example — let us know and we'll add it to the roadmap.
Is it possible to disable access to server logs and limit the integration to crawl data and Google Search Console only?
Yes. Let us know and we'll apply it to your account — you'll receive confirmation once it's done.
Are audit logs available for MCP activity, including queries
and actions performed
by the AI agent?
We don't see your AI conversations. On our side, we log the MCP tool requests made against
your account — the tool calls (query, count, start_crawl, etc.) — not the chat content.
What data is transmitted
to the AI model when using
this integration?
Only what the AI requests to answer your prompt: crawl data, GSC data, log columns and URLs from your account can be returned
to your AI client (Claude, ChatGPT, etc.) on demand. JetOctopus never sends your data to an AI provider on its own. The AI client you connect makes the calls and your data is not used to train models.
Can the integration be restricted to specific users or groups within the organization?
Not available today, but we can implement it on request.

Your crawl data, your logs, your GSC. All in one chat

Connect JetOctopus MCP