How to Perform a Full GEO + SEO Audit with Claude Code CLI

The geo-seo-claude repository enables a complete Generative Engine Optimization (GEO) audit through a single /geo audit command in Claude Code CLI, crawling up to 50 pages and aggregating six weighted category scores into a composite 0-100 GEO score.

Performing a comprehensive GEO + SEO audit requires analyzing how AI engines perceive, cite, and rank your content across platforms like ChatGPT, Perplexity, and Google AI Overviews. The zubair-trabzada/geo-seo-claude open-source project implements this as a native Claude Code CLI skill, orchestrating parallel sub-agents that evaluate technical infrastructure, content authority, and platform readiness.

The Three-Phase Audit Architecture

The audit workflow defined in skills/geo-audit/SKILL.md executes through three distinct phases that transform raw website data into actionable intelligence.

Phase 1: Discovery and Crawling

When you invoke the audit command, the CLI initiates Discovery by fetching the homepage to determine the business type, then crawls up to 50 pages via sitemap or internal link analysis. This phase relies on scripts/fetch_page.py to retrieve and parse HTML content, establishing the dataset for downstream analysis.

Phase 2: Parallel Sub-Agent Analysis

Five specialized sub-agents run concurrently to evaluate distinct optimization vectors:

• AI Visibility (agents/geo-ai-visibility.md): Assesses citability, crawler access permissions, and brand authority metrics using scripts/citability_scorer.py and scripts/brand_scanner.py.
• Platform Optimization (agents/geo-platform-analysis.md): Evaluates readiness for ChatGPT, Perplexity, Google AI Overviews, Gemini, and Bing Copilot.
• Technical SEO (agents/geo-technical.md): Audits robots.txt, server-side rendering (SSR), Core Web Vitals, security headers, and mobile responsiveness.
• Content E-E-A-T (agents/geo-content.md): Scores Experience, Expertise, Authoritativeness, and Trustworthiness signals.
• Schema Markup (agents/geo-schema.md): Detects, validates, and generates JSON-LD structured data using templates from the schema/ directory.

Each sub-agent produces a 0-100 score and a detailed findings report.

Phase 3: Score Aggregation and Reporting

The composite GEO score calculates as a weighted average: citability 25%, brand 20%, E-E-A-T 20%, technical 15%, schema 10%, and platform 10%. The orchestrator generates GEO-AUDIT-REPORT.md containing severity-ranked issues and a prioritized 30-day action plan.

How to Run a Full GEO Audit from the CLI

Execute the complete audit workflow with the primary command:

/geo audit https://example.com

This command coordinates the three-phase pipeline: discovery, parallel sub-agent analysis, and weighted score aggregation. The output writes to GEO-AUDIT-REPORT.md in your current working directory, containing technical diagnostics and remediation steps.

Generating Client-Ready Reports

Transform raw audit data into polished deliverables for stakeholders:

/geo report https://example.com

This generates GEO-CLIENT-REPORT.md with an executive summary, score dashboard, and prioritized action plan.

For PDF distribution (requires the reportlab Python package):

/geo report-pdf https://example.com

This calls scripts/generate_pdf_report.py to produce GEO-REPORT-<brand>.pdf with color-coded score gauges, methodology appendices, and visual charts rendered from templates/geo-report-template.html and templates/geo-report-style.css.

Quick Audits and Individual Sub-Skills

For rapid assessment before committing to a full crawl:

/geo quick https://example.com

Returns an inline terminal summary and GEO score within 60 seconds, optionally saving to a CRM prospect record.

Run individual analysis modules by calling specific sub-skills:

/geo technical https://example.com

This invokes only the technical SEO agent (agents/geo-technical.md), outputting GEO-TECHNICAL-AUDIT.md with detailed infrastructure checklists.

Core Components and File Structure

The audit system relies on these key files within the repository:

Component	Role	Key Files
Skill Router	Routes /geo commands to sub-skills	geo/SKILL.md
Audit Orchestrator	Implements the three-phase workflow	skills/geo-audit/SKILL.md
Sub-Agents	Domain-specific analysis modules	agents/geo-ai-visibility.md, agents/geo-platform-analysis.md, agents/geo-technical.md, agents/geo-content.md, agents/geo-schema.md
Python Utilities	Data fetching and scoring logic	scripts/fetch_page.py, scripts/citability_scorer.py, scripts/brand_scanner.py, scripts/llmstxt_generator.py, scripts/generate_pdf_report.py
Schema Templates	JSON-LD examples for validation	schema/*.json
Report Templates	PDF generation assets	templates/geo-report-template.html, templates/geo-report-style.css

Reference these files in docs/architecture.md and docs/commands-reference.md for deep technical implementation details.

Summary

• Single-command execution: Run /geo audit <url> to trigger the complete three-phase workflow.
• Parallel analysis: Five specialized sub-agents evaluate AI visibility, platform readiness, technical SEO, E-E-A-T, and schema markup simultaneously.
• Weighted scoring: The composite GEO score applies specific weights (citability 25%, brand 20%, E-E-A-T 20%) to prioritize AI-citation factors.
• Multiple output formats: Generate Markdown technical reports, client-ready summaries, or styled PDFs via /geo report and /geo report-pdf.
• Modular architecture: Individual sub-skills like /geo technical allow targeted audits without full site crawls.

Frequently Asked Questions

How long does a full GEO audit take to complete?

A standard audit crawls up to 50 pages and typically completes in 3-5 minutes depending on site speed and complexity. The /geo quick command provides a 60-second snapshot for rapid assessment, while full PDF generation adds additional processing time for rendering charts and formatting.

What is the difference between GEO and traditional SEO?

Generative Engine Optimization focuses specifically on how AI systems (ChatGPT, Perplexity, Google AI Overviews) cite, summarize, and recommend content, whereas traditional SEO targets ranking in conventional search results. The audit weights citability (25%) and platform optimization (10%) heavily because these metrics directly influence AI citation behavior rather than just SERP positioning.

Can I customize the audit crawl depth or scoring weights?

The default configuration in skills/geo-audit/SKILL.md limits crawling to 50 pages to balance thoroughness with execution time. While the weighted scoring formula (citability 25%, brand 20%, etc.) is hardcoded in the current implementation, you can run individual sub-skills like /geo technical or /geo content to focus analysis on specific categories without the composite score aggregation.

What files are generated after running the audit?

The primary output is GEO-AUDIT-REPORT.md containing technical findings and action plans. When using /geo report, the system creates GEO-CLIENT-REPORT.md for stakeholder presentations. The /geo report-pdf command generates GEO-REPORT-<brand>.pdf using the HTML/CSS templates in the templates/ directory and the reportlab Python library for professional distribution.

{
  "mcpServers": {
    "instagit": {
      "command": "npx",
      "args": ["-y", "instagit@latest"]
    }
  }
}