agentleFS
forefyforefy/.context/skills/generic-utils/agentic-seo/SKILL.md

agentic-seo

Skill153 starsChanged 44 days agoMITView on GitHub

Turn a public-facing website into optimized for search engines and generative search engines by using all efficient means

2 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

9 rules. About 0 are common, 8 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈8

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use the Open Graph checker https://www.opengraph.xyz/ to ensure that when sites are sent over social media, Open Graph is optimized per …

    line 10
  2. Distinctive:

    Run a generic security audit because stuff like immature security headers revealing technologies, stack traces, and console log errors may …

    line 12
  3. Distinctive:

    Check for llms.txt - and concisely create one if missing.

    line 13
  4. Distinctive:

    Ensure correctly generating sitemap.xml (with entries that won’t lower our score)

    line 14
  5. Distinctive:

    Use a browser driver to go to https://search.google.com/ and ensure the site is listed on Google Search Console as an asset, the sitemap is …

    line 15
  6. Distinctive:

    Ensure accessibility, privacy, terms, are applied to high extend (unless the user opts out of privacy/terms) to best results

    line 17
  7. Very specific:

    If the site is on Cloudflare, turn on Crawler Hints, otherwise, use IndexNow

    line 19
  8. Distinctive:

    Check for optimized LCP (Largest Contentful Paint) and FCP (First Contentful Paint)

    line 20
  9. Distinctive:

    Always cross-check optimizations to see we aren’t breaking anything.

    line 26

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Go over the source code and live site (ask the user) and optimize GEO and SEO according to obvious judgement, but don’t leave out any of these:

  • Go to https://schema.org/docs/schemas.html and browse schemas we should use, as well as breadcrumbs, for rich results and JSON-LD content for search engines
  • Drive a browser to https://www.seoptimer.com/ and run a free login-less site audit and apply all code adjustments to satisfy and get a top score (has 3 attempts a day, alternative fallback to https://www.seobility.net/en/seocheck)
  • Use the Open Graph checker https://www.opengraph.xyz/ to ensure that when sites are sent over social media, Open Graph is optimized per page or important page we are sharing in-app.
  • User might have a CDN like Cloudflare that does caching, or any other infrastructural service that we don’t want to conflict with; we can use the browser and ask the user to log in there too for tweaks.
  • Run a generic security audit because stuff like immature security headers revealing technologies, stack traces, and console log errors may lower SEO
  • Check for llms.txt - and concisely create one if missing.
  • Ensure correctly generating sitemap.xml (with entries that won’t lower our score)
  • Use a browser driver to go to https://search.google.com/ and ensure the site is listed on Google Search Console as an asset, the sitemap is submitted lately, and inspect any page issues and fix them.
  • Adapt keywords per page for the content and highly optimize for specific search queries (best if you consult with the user on what you think is the right set)
  • Ensure accessibility, privacy, terms, are applied to high extend (unless the user opts out of privacy/terms) to best results
  • Also use LightHouse for page speed insights https://pagespeed.web.dev/analysis
  • If the site is on Cloudflare, turn on Crawler Hints, otherwise, use IndexNow
  • Check for optimized LCP (Largest Contentful Paint) and FCP (First Contentful Paint)

Limits to realize

  • Using harness's native browsers sometimes easily give us off as bots, so we might want to involve the user in using his native browser to his choice, and click any captcha screen verifications or etc.
  • We don’t want to break anything or over-optimize, especially in non-significant, riskier areas.
  • Always cross-check optimizations to see we aren’t breaking anything.
  • Don’t stray into creating content or feature additions that might improve SEO if the user hadn’t asked for it- but you may very well suggest those.
  • Some site warnings are on code we can't control (like analytics 3rd parties, cdn-added scripts etc - we can ignore those.

Verification

Your goal is only done after site with all fixes went to prod, we verified everything is healthy, and seoptimer, opengraph, google search and all the tools we used the browser to optimize are now close to perfect or as best as we can achieve without sacrificing app logic or user intents.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.