← all repositories
zubair-trabzada/geo-seo-claude

Auditing websites for GPTBot, not just Googlebot

It adds a GEO audit suite to Claude Code that scores how easily AI search engines can find and cite your site, since traditional SEO misses the LLM crawler entirely.

★11.2k stars Python Other AI
geo-seo-claude
Velocity · 7d
+25
★ / day
Trend
↗accelerating
star history

What it does geo-seo-claude is a skill for the Claude Code CLI that runs comprehensive website audits targeting AI search engines—what the author calls Generative Engine Optimization (GEO). It fetches a homepage, detects business type, and spins up five parallel subagents to check AI citability, platform-specific readiness, technical SEO, content quality, and schema markup. The output is a composite GEO Score from 0 to 100 and a prioritized action plan, optionally rendered as a client-ready PDF report with charts and score gauges.

The interesting bit The project treats AI crawlers as a first-class audience: it checks robots.txt against 14+ LLM bots like GPTBot and ClaudeBot, scans brand mentions across Reddit and Wikipedia, and even generates an llms.txt file to help crawlers map your site. It is essentially a Python-backed orchestration layer that uses Claude Code’s own agent system to parallelize work that would otherwise require a stack of separate SaaS tools.

Key highlights

  • Runs inside Claude Code as a namespaced skill, extending the CLI with audit and reporting capabilities.
  • Parallel audit flow using five subagents to analyze AI visibility, platforms, technical SEO, content, and schema simultaneously.
  • Scores content blocks for “citability,” claiming optimal AI-cited passages run 134–167 words and should be self-contained.
  • Includes a lightweight CRM and proposal generator for agencies, storing prospect data in ~/.geo-prospects/.
  • Ships with JSON-LD schema templates (Organization, LocalBusiness, Product, etc.) and a ReportLab-based PDF report engine.

Caveats

  • Requires the Claude Code CLI and Python 3.8+; it is not a standalone application.
  • On Windows, the installer expects Git Bash and explicitly warns against PowerShell or Command Prompt.
  • The prospect data directory (~/.geo-prospects/) is left behind after uninstall and must be deleted manually.

Verdict Worth a look if you run an agency or marketing team that needs to explain AI-search readiness to clients with hard numbers and pretty PDFs. Skip it if you are not already using Claude Code or if you need a fully self-hosted, independent audit tool.

Frequently asked

What is zubair-trabzada/geo-seo-claude?
It adds a GEO audit suite to Claude Code that scores how easily AI search engines can find and cite your site, since traditional SEO misses the LLM crawler entirely.
Is geo-seo-claude open source?
Yes — zubair-trabzada/geo-seo-claude is open source, released under the MIT license.
What language is geo-seo-claude written in?
zubair-trabzada/geo-seo-claude is primarily written in Python.
How popular is geo-seo-claude?
zubair-trabzada/geo-seo-claude has 11.2k stars on GitHub and is currently accelerating.
Where can I find geo-seo-claude?
zubair-trabzada/geo-seo-claude is on GitHub at https://github.com/zubair-trabzada/geo-seo-claude.

heatdrop uses Google Analytics to see which pages get read — nothing else. Your call. How we handle data.