auditing-seo
Skill0 starsChanged 4 months agoMITView on GitHub
Audits a website's technical and on-page SEO - crawlability, metadata, and heading structure - checked against real fetched page content, not a generic checklist. Use whenever the user wants an SEO audit, wants to know why a page isn't ranking, or asks for SEO recommendations for a site.
In this repository19 other files it gives its agents.
- adding-authenticationSkill
- automating-git-workflowsSkill
- building-email-sequencesSkill
- building-graphql-apisSkill
- building-wordpress-pluginsSkill
- designing-carouselsSkill
- fine-tuning-ml-modelsSkill
- handling-web-formsSkill
- n8n-automationSkill
- organizing-spreadsheetsSkill
- reviewing-cloudflare-workers-codeSkill
- reviewing-code-for-securitySkill
- scaffolding-landing-pagesSkill
- setting-up-error-monitoringSkill
- setting-up-firebase-authSkill
- using-nextjs-best-practicesSkill
- working-with-postgres-databasesSkill
- writing-marketing-copySkill
- writing-terraform-codeSkill
The file
We split this skill into the rules it gives an agent
2 rules. About 0 are common, 2 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 12
Fetch the actual page: before auditing anything - title, meta description, canonical, and heading structure need to be read from the real …
- Distinctive: line 39
Don't recommend keyword stuffing or any tactic that trades short-term ranking for penalty risk - flag black-hat-adjacent requests and …
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Auditing SEO
Audits technical and on-page SEO by actually fetching the page and checking real markup, not describing what a hypothetical page might have wrong.
Workflow
- Fetch the actual page before auditing anything - title, meta description, canonical, and heading structure need to be read from the real response, not assumed from the URL or a description of the site.
- Technical foundation first: crawlability (robots.txt, sitemap.xml), indexability (no accidental
noindex), Core Web Vitals. - On-page basics per key page: title tag, meta description, exactly one H1, logical heading hierarchy below it - see the real finding below on why "exactly one" matters in practice, not just in theory.
- Content-search intent match: does the content actually answer what the target query is asking.
- Prioritize the fix list by effort vs impact, not just severity.
Real audit example: stripe.com/blog
Fetched the live page and checked it against the on-page basics - here's what a real audit of a real, well-resourced site's page actually turns up.
Title tag: Stripe Blog: Online Payment Solutions Blog - 43 characters, descriptive, well within the ~60-character display limit. Pass.
Meta description: Follow the Stripe blog to learn about new product features, the latest in technology, payment solutions, and business initiatives. - 134 characters, within the ~155-character guideline, accurately describes the page. Pass.
Canonical tag: present, self-referencing (https://stripe.com/blog). Pass.
Open Graph / Twitter tags: og:title, og:description, og:image, twitter:card, twitter:image all populated - social shares will render correctly. Pass.
H1 structure - real issue found: the page markup shows dozens of H1-level headings, not one. Every mega-menu category (# Payments, # Revenue, # Money Management, # Platforms and marketplaces, # More) and every single blog post preview title on the index page (# Four travel and hospitality trends from HITEC 2026, # What Link data tells us about AI spending, # Stripe Projects adds new agent integrations..., and so on through the full list of ~15 preview cards) is marked as an H1, alongside recurring # Recent highlights and individual highlight titles.
This is the textbook version of the "exactly one H1" rule from the checklist above, found on a real, professionally-built page: with dozens of H1s, there's no single clear signal to search engines about what this page is fundamentally about - the semantic weight that should concentrate on "Stripe Blog" as the page's own H1 is instead diluted across every nav category and every post preview card.
Practical audit note this produces: don't take "the site looks polished" as a proxy for "the on-page SEO is clean" - a large, professionally maintained site can still carry a structural heading issue like this, likely from a CMS template that maps every card/module title to an <h1> by default rather than an appropriately nested <h2>/<h3>. This is exactly the kind of finding that's invisible from reading rendered content and only shows up by actually checking the markup.
Notes
- Heading levels above were read from the page's actual structure via a real fetch, not inferred from visual appearance - visual heading size on a page doesn't reliably indicate the underlying tag level, which is what search engines parse.
- Don't recommend keyword stuffing or any tactic that trades short-term ranking for penalty risk - flag black-hat-adjacent requests and suggest the sustainable alternative instead.
- Prioritize fixes by effort vs impact: in the example above, the H1 issue is a template-level fix (one change, applies sitewide) despite looking like "dozens of issues" when counted per-page - flag it as one fix with wide impact, not dozens of separate findings.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

