Claude SEO is an open-source SEO audit plugin for Claude Code, created by Agrici Daniel and released under the MIT license; it is an independent community project, not an Anthropic product. Given a URL, it detects the type of site (SaaS, local business, e-commerce, publisher or agency), adjusts its checks and fans the audit out to specialist agents for technical SEO, content, schema, performance, AI search and other areas, with up to 17 working in parallel. The result is a scored, prioritized action plan in which each recommendation states what it is based on and how to check that the fix worked.
The plugin bundles 26 sub-skills and 19 agents behind a single /seo command. Commands cover full-site and single-page audits, crawlability, sitemaps, hreflang, images, content briefs and E-E-A-T analysis, Schema.org detection, validation and generation, Generative Engine Optimization checks for Google AI Overviews, ChatGPT, Perplexity and Gemini, an agent-readiness check, local SEO and Google Business Profile, backlink profile analysis from Moz, Bing and Common Crawl data, and change monitoring. JavaScript-heavy pages are rendered in a headless browser.
Claude SEO installs from its own Claude Code plugin marketplace or with a manual script on macOS, Linux and Windows, and /seo setup creates an isolated Python environment and headless Chromium. It needs Claude Code 1.0.33 or newer and Python 3.11 to 3.14. Pages and data it analyzes are sent to Claude like any other Claude Code work, while reports stay on your machine as Markdown, with PDF output available; free Google APIs can be connected for field data, and optional extensions link paid services such as SE Ranking and Profound with your own accounts.
The tool itself is free. Costs come from the Claude Code plan it runs on: five agents that need more judgment use Opus and the rest use Sonnet, so a full audit uses more of a plan than a single-page check, and each agent's model can be changed. Without Google API keys, Core Web Vitals are lab estimates and indexing is inferred. The author also offers a private version with early features to members of his paid AI Marketing Hub Pro community; the public repository remains MIT.
