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.
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 9.1k stars on GitHub and is currently cooling off.
- Where can I find geo-seo-claude?
- zubair-trabzada/geo-seo-claude is on GitHub at https://github.com/zubair-trabzada/geo-seo-claude.