Reference tags tell AI systems where to find your structured files. Without them, the files exist but are invisible to crawlers.
What to add
Add <link rel="alternate"> tags inside the <head> of each page, pointing to the matching Geordy files.
Homepage (8 files)
The homepage includes llms.txt and humans.txt — root-only files. Note that llms.txt uses rel="llms" per the llmstxt.org spec.
Homepage <head>HTML
<!-- Geordy AI References -->
<link rel="llms" type="text/plain" href="https://ai.yourdomain.com/llms.txt" title="LLMs.txt" data-ai="true" />
<link rel="alternate" type="text/plain" href="https://ai.yourdomain.com/humans.txt" title="Humans.txt" data-ai="true" />
<link rel="alternate" type="text/markdown" href="https://ai.yourdomain.com/index.md" title="Markdown" data-ai="true" />
<link rel="alternate" type="application/ld+json" href="https://ai.yourdomain.com/index.schema.json" title="Schema" data-ai="true" />
<link rel="alternate" type="application/yaml" href="https://ai.yourdomain.com/index.yaml" title="YAML" data-ai="true" />
<link rel="alternate" type="application/json" href="https://ai.yourdomain.com/index.og.json" title="Open Graph" data-ai="true" />
<link rel="alternate" type="application/rss+xml" href="https://ai.yourdomain.com/index.xml" title="RSS Feed" data-ai="true" />
<link rel="alternate" type="application/json" href="https://ai.yourdomain.com/index.manifest.json" title="Manifest" data-ai="true" />Inner pages (6 files)
Inner pages use the page path as the filename, preserving directory structure. Root-only files (llms.txt, humans.txt) can still be included — they always point to the same root URL.
/about/team page <head>HTML
<link rel="alternate" type="text/markdown" href="https://ai.yourdomain.com/about/team.md" title="Markdown" data-ai="true" />
<link rel="alternate" type="application/ld+json" href="https://ai.yourdomain.com/about/team.schema.json" title="Schema" data-ai="true" />
<link rel="alternate" type="application/yaml" href="https://ai.yourdomain.com/about/team.yaml" title="YAML" data-ai="true" />
<link rel="alternate" type="application/json" href="https://ai.yourdomain.com/about/team.og.json" title="Open Graph" data-ai="true" />
<link rel="alternate" type="application/rss+xml" href="https://ai.yourdomain.com/about/team.xml" title="RSS Feed" data-ai="true" />
<link rel="alternate" type="application/json" href="https://ai.yourdomain.com/about/team.manifest.json" title="Manifest" data-ai="true" />Path mapping: /about → about.*, /pricing → pricing.*, /about/team → about/team.*
Dynamic implementation (recommended)
Use a reusable React component that automatically generates correct file URLs based on the current page path. This ensures every page points to its own Geordy files without manual updates.
components/geordy-tags.tsxTSX
import React from "react"
interface GeordyTagsProps {
domain: string
pagePath?: string
formats?: {
yaml?: boolean; markdown?: boolean; llms?: boolean; schema?: boolean
rss?: boolean; manifest?: boolean; humans?: boolean; og?: boolean
}
}
export function GeordyTags({ domain, pagePath = "/", formats = {} }: GeordyTagsProps) {
const today = new Date().toISOString().slice(0, 10)
const isRoot = pagePath === "/"
// Preserve path structure: /about/team -> about/team
const pathSegment = isRoot ? "index" : pagePath.replace(/^\//, "")
const base = `https://files.geordy.ai/${domain}`
// Default all formats to true
const { yaml = true, markdown = true, llms = true, schema = true, rss = true, manifest = true, humans = true, og = true } = formats
return (
<>
{/* Per-page formats */}
{yaml && <link rel="alternate" type="application/yaml" href={`${base}/${pathSegment}.yaml`} title="YAML" data-ai="true" />}
{markdown && <link rel="alternate" type="text/markdown" href={`${base}/${pathSegment}.md`} title="Markdown" data-ai="true" />}
{schema && <link rel="alternate" type="application/ld+json" href={`${base}/${pathSegment}.schema.json`} title="Schema" data-ai="true" />}
{rss && <link rel="alternate" type="application/rss+xml" href={`${base}/${pathSegment}.xml`} title="RSS Feed" data-ai="true" />}
{manifest && <link rel="alternate" type="application/json" href={`${base}/${pathSegment}.manifest.json`} title="Manifest" data-ai="true" />}
{og && <link rel="alternate" type="application/json" href={`${base}/${pathSegment}.og.json`} title="Open Graph" data-ai="true" />}
{/* Root-only formats */}
{llms && <link rel="llms" href={`${base}/llms.txt`} title="LLMs.txt" data-ai="true" data-version={today} />}
{humans && <link rel="alternate" type="text/plain" href={`${base}/humans.txt`} title="Humans.txt" data-ai="true" />}
</>
)
}Usage in Next.js
app/layout.tsxTSX
"use client"
import { usePathname } from "next/navigation"
import { GeordyTags } from "@/components/geordy-tags"
return function Layout({ children }) {
const pathname = usePathname()
return (
<html>
<head>
<GeordyTags domain="yourdomain.com" pagePath={pathname} />
</head>
<body>{children}</body>
</html>
)
}
// To disable specific formats:
<GeordyTags domain="yourdomain.com" pagePath={pathname} formats={{ rss: false, manifest: false }} />Benefits: Tags auto-generate for every page. New pages are covered automatically. No manual updates needed.
Verification
- 1View page source and search for your subdomain — tags should appear in the <head>
- 2Click one of the href URLs — the file should load
- 3Check multiple pages to confirm dynamic implementation works
This is the activation step. Files that exist but aren't referenced are invisible to AI systems.