Diagnose AI search readiness and generate client reports
See whether your website is mentioned in AI search engines like ChatGPT or Perplexity and understand why. This tool analyzes your site to score its readiness for AI citation, identify technical issues, and produce professional reports for clients. It runs as a skill in the Claude Code environment, handling complex SEO tasks with a single command.
Why respond to the surge in AI search traffic
Traditional search traffic is declining while AI-driven inflow is growing rapidly. According to the README, AI referral traffic has increased by 527% year-over-year, and AI traffic conversion rates are 4.4 times higher than standard organic traffic. Additionally, brand mentions in AI search show a correlation three times stronger than backlinks. This tool focuses on optimizing websites to match these changes.
The diagnosis and report generation process
Users initiate a full diagnosis by entering the /geo audit <url> command. The system fetches the homepage, detects the business type, and then runs five sub-agents simultaneously. Each agent handles AI visibility, platform analysis, technical SEO, content quality, and schema markup. Once analysis is complete, a comprehensive GEO score from 0 to 100 is calculated, and a prioritized action plan is generated. The final output is delivered as a Markdown report or a PDF file containing charts and visualizations.
Key analysis items and features
The tool provides citability scoring to evaluate AI citation readiness.
- Citability scoring Analyzes whether content blocks are suitable for AI citation, scoring them based on optimal sentence length and structure.
- AI crawler analysis Inspects the
robots.txtfile to verify access permissions for major AI crawlers and recommends specific configurations. - Brand mention scan Tracks how the brand is mentioned on key platforms where AI citations occur.
- llms.txt generation Analyzes or generates
llms.txt, an emerging standard file that helps AI crawlers understand site structure.
Installation and requirements
Installation on macOS and Linux is done with a single curl command, while Windows users install via Git Bash. Python 3.8 or higher, Claude Code CLI, and Git are required. uv and Playwright are optional, with the latter used for screenshot generation. Python dependencies are installed in a dedicated virtual environment at ~/.claude/skills/geo/.venv/, so they do not affect the system Python. Installation is handled automatically by running the install.sh script.
Suitable audiences and considerations
This tool is suitable for GEO agencies, marketing teams, content creators, local businesses, SaaS companies, e-commerce sites, and any website needing improved AI search visibility. It is particularly useful for agencies or marketing personnel who must deliver reports to clients. However, the tool focuses on diagnosis and report generation; actual content modifications or technical implementation must be performed by the user. Additionally, CRM and proposal-related data are stored in the ~/.geo-prospects/ directory and are not automatically deleted upon uninstallation, requiring manual cleanup. It is distributed under the MIT license, and the community offers ways to learn about commercialization.