Built at work · description only In progress

SEO Potential Analysis with MCP

Connecting Screaming Frog, Google Search Console and Sistrix to Claude Code, so crawl, ranking and click data can be analysed together.

Problem

Crawl, ranking and click data sat in three tools with three separate exports.

Solution

Screaming Frog, Google Search Console and Sistrix are connected to Claude Code via MCP (Model Context Protocol) and OAuth. Reusable Claude Skills define the standard analyses and how recommended actions are derived from the data, so an analysis runs the same way each time.

Architecture of the SEO potential analysis Screaming Frog crawl data, Google Search Console click data and Sistrix ranking data connect via MCP servers and OAuth to Claude Code. Reusable Claude Skills guide the analysis. The output is an analysis with recommended actions. Screaming Frog crawl data Search Console click data Sistrix ranking data MCP servers · OAuth Claude Code Claude Skills standard analyses Analysis + recommended actions
Schematic overview of the setup in progress. No internal interfaces or data shown.

Stack

  • MCP (Model Context Protocol)
  • OAuth
  • Claude Code
  • Claude Skills
  • Screaming Frog
  • Google Search Console
  • Sistrix

Result

In progress.

No results to report yet.

My role

Building the setup: the MCP connections and the Claude Skills.

What I learned

The hard part isn't connecting the tools, it's deciding what a good analysis looks like before the data arrives. Reusable Claude Skills turned that into a repeatable standard instead of a fresh judgment call every time.

Back to all projects