The latest articles on DEV Community by LIUBOJUN (@liubojun). https://dev.to/liubojun https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4181309%2F45593f12-b567-4fc3-96ae-9ea147223029.jpg https://dev.to/liubojun en LIUBOJUN Tue, 13 Oct 2026 06:59:30 +0000 https://dev.to/liubojun/i-open-sourced-the-seo-audit-i-run-every-day-one-command-zero-dependencies-e9a https://dev.to/liubojun/i-open-sourced-the-seo-audit-i-run-every-day-one-command-zero-dependencies-e9a <p>SEO tooling is in an awkward place: either a monthly fee that isn't cheap, or free but only skin-deep. I run site audits every day inside my own AI operations system for e-commerce stores, so I packaged the whole routine into a tool and open-sourced it.</p> <p>It runs as a plain CLI <strong>or</strong> as a skill for your AI coding agent (Claude Code, Codex, opencode…).<br> </p> <div class="highlight js-code-highlight"> <pre class="highlight shell"><code>git clone https://github.com/zxhydfzr/website-seo-audit.git <span class="nb">cd </span>website-seo-audit python3 <span class="nt">-m</span> seo_audit https://yoursite.com </code></pre> </div> <p>No API keys, no sign-up, no dependencies. Just Python 3.8+ and the standard library. It's also on PyPI (<code>pip install website-seo-audit</code>).</p> <h2> What it checks: three layers </h2> <p><strong>Page layer.</strong> Title, meta description, heading hierarchy, image alt text, internal linking. Fundamentals aren't glamorous, but they're exactly where most sites' ranking problems live.</p> <p><strong>Technical layer.</strong> robots.txt, sitemap, canonical tags, HTTPS, mobile handling: the site-wide settings that decide whether pages get crawled and indexed at all. Break this layer and it doesn't matter how good the writing is.</p> <p><strong>Structured data layer.</strong> Completeness and correctness of JSON-LD. I deliberately made this part heavier. Structured data doesn't just drive Google's rich results; it also affects whether AI search (ChatGPT, Perplexity and the rest) can understand and cite your pages correctly, which is what people now call GEO.</p> <h2> Answers, not homework </h2> <p>Each run produces a scored report: a grade per layer, issues ranked by severity, and a concrete fix attached to every issue.<br> </p> <div class="highlight js-code-highlight"> <pre class="highlight plaintext"><code>Score: 72/100 (C) · 🔴 2 critical · 🟡 8 warnings · 🟢 15 notices Crawled 23 pages · 45 internal links checked 🔴 Broken internal links (3×) Fix: Fix or remove the link, or restore the target page (add a redirect if it moved). 🟡 Structured data missing fields: Article missing image, datePublished </code></pre> </div> <p>The goal is that you finish reading and start fixing, instead of getting a pile of data to interpret.</p> <h2> Use it inside your AI agent </h2> <p>The repo ships a <code>SKILL.md</code>, so for Claude Code it's one line:<br> </p> <div class="highlight js-code-highlight"> <pre class="highlight shell"><code>git clone https://github.com/zxhydfzr/website-seo-audit ~/.claude/skills/website-seo-audit </code></pre> </div> <p>Then just say <em>"audit the SEO of <a href="https://example.com" rel="noopener noreferrer">https://example.com</a>"</em> and the agent runs the audit and walks you through the results.</p> <h2> What it deliberately doesn't do </h2> <p>No performance testing. Lighthouse is already the standard answer there, and running both gives you a complete check-up.</p> <p>Repo: <a href="https://github.com/zxhydfzr/website-seo-audit" rel="noopener noreferrer">https://github.com/zxhydfzr/website-seo-audit</a> (stars and issues welcome)</p> <p><em>Originally published on <a href="https://www.liubojun.com/en/seo-audit-skill-open-source/" rel="noopener noreferrer">Chill AI Automate</a>, where I also keep a <a href="https://www.liubojun.com/en/skills/" rel="noopener noreferrer">list of open-source SEO tools I've vetted</a>.</em></p> opensource ai seo