fix(seo): make content pagination indexable (#8651)

Author: waleedlatif1

.claude/rules/landing-seo-geo.md

@@ -17,6 +17,7 @@ paths:
 - All copy is server-rendered text: no text baked into images, no content that exists only after a client effect runs.
 - Navbar is a Server Component (no `'use client'`) for immediate crawlability. Logo `<Image>` has `priority` (LCP element). The navbar `<nav>` carries `SiteNavigationElement` schema.org markup.
 - Structured data: emit JSON-LD (`Organization`, `WebSite`, `WebApplication` with `featureList`, `FAQPage` if an FAQ exists) from a server component rendered before visible content. Keep `featureList` in sync with the features the page shows (`apps/sim/app/(landing)/components/home-structured-data/`).
+- `/library` and `/blog` pagination stays server-rendered with real next/previous links. Each unfiltered page is indexable and uses its own canonical URL (`?page=2`, etc.); page 1 canonicalizes to the bare index. Tag-filtered variants remain `noindex,follow`. See [Google pagination guidance](https://developers.google.com/search/docs/specialty/ecommerce/pagination-and-incremental-page-loading).
 - After adding routes or anchors, verify `app/sitemap.ts` and `app/robots.ts` still reflect reality.
 
 ## GEO (Generative Engine Optimisation)

.cursor/rules/landing-seo-geo.mdc

@@ -17,6 +17,7 @@ globs: ["apps/sim/app/(landing)/**/*.tsx","apps/sim/content/**/*.mdx"]
 - All copy is server-rendered text: no text baked into images, no content that exists only after a client effect runs.
 - Navbar is a Server Component (no `'use client'`) for immediate crawlability. Logo `<Image>` has `priority` (LCP element). The navbar `<nav>` carries `SiteNavigationElement` schema.org markup.
 - Structured data: emit JSON-LD (`Organization`, `WebSite`, `WebApplication` with `featureList`, `FAQPage` if an FAQ exists) from a server component rendered before visible content. Keep `featureList` in sync with the features the page shows (`apps/sim/app/(landing)/components/home-structured-data/`).
+- `/library` and `/blog` pagination stays server-rendered with real next/previous links. Each unfiltered page is indexable and uses its own canonical URL (`?page=2`, etc.); page 1 canonicalizes to the bare index. Tag-filtered variants remain `noindex,follow`. See [Google pagination guidance](https://developers.google.com/search/docs/specialty/ecommerce/pagination-and-incremental-page-loading).
 - After adding routes or anchors, verify `app/sitemap.ts` and `app/robots.ts` still reflect reality.
 
 ## GEO (Generative Engine Optimisation)

apps/sim/app/(landing)/blog/page.tsx

@@ -1,21 +1,19 @@
 import type { Metadata } from 'next'
+import { notFound } from 'next/navigation'
 import { getAllPostMeta } from '@/lib/blog/registry'
 import { BLOG_SECTION, buildCollectionPageJsonLd, buildIndexMetadata } from '@/lib/blog/seo'
-import { selectVisiblePosts } from '@/lib/content/index-list'
+import { resolveContentPage, selectVisiblePosts } from '@/lib/content/index-list'
 import { ContentIndexPage } from '@/app/(landing)/components'
 
-/**
- * Filtered/paginated variants render genuinely different lists, but only the
- * bare index is indexable — see `buildIndexMetadata` in `@/lib/content/seo`
- * for the shared noindex policy.
- */
 export async function generateMetadata({
   searchParams,
 }: {
   searchParams: Promise<{ page?: string; tag?: string }>
 }): Promise<Metadata> {
   const { page, tag } = await searchParams
-  const pageNum = Math.max(1, Number(page || 1))
+  const posts = await getAllPostMeta()
+  const pageNum = resolveContentPage(posts, { page, tag })
+  if (pageNum === null) notFound()
   return buildIndexMetadata({ tag, pageNum })
 }
 
@@ -25,8 +23,9 @@ export default async function BlogIndex({
   searchParams: Promise<{ page?: string; tag?: string }>
 }) {
   const { page, tag } = await searchParams
-  const pageNum = Math.max(1, Number(page || 1))
   const posts = await getAllPostMeta()
+  const pageNum = resolveContentPage(posts, { page, tag })
+  if (pageNum === null) notFound()
 
   return (
     <ContentIndexPage

apps/sim/app/(landing)/components/content-index-page/content-index-page.tsx

@@ -22,8 +22,8 @@ interface ContentIndexPageProps {
  * Shared index/list layout for a content section (blog or library): section
  * header, featured-post row, remaining posts list, and pagination. Both
  * sections render this exact layout, parameterized by `basePath` and copy —
- * see `.claude/rules/landing-seo-geo.md` for the filtered/paginated noindex
- * policy this pairs with (`buildIndexMetadata` in `@/lib/content/seo`).
+ * paginated URLs are indexable, while tag-filtered variants are noindexed
+ * by `buildIndexMetadata` in `@/lib/content/seo`.
  */
 export function ContentIndexPage({
   basePath,

apps/sim/app/(landing)/library/page.tsx

@@ -1,21 +1,19 @@
 import type { Metadata } from 'next'
-import { selectVisiblePosts } from '@/lib/content/index-list'
+import { notFound } from 'next/navigation'
+import { resolveContentPage, selectVisiblePosts } from '@/lib/content/index-list'
 import { getAllPostMeta } from '@/lib/library/registry'
 import { buildCollectionPageJsonLd, buildIndexMetadata, LIBRARY_SECTION } from '@/lib/library/seo'
 import { ContentIndexPage } from '@/app/(landing)/components'
 
-/**
- * Filtered/paginated variants render genuinely different lists, but only the
- * bare index is indexable — see `buildIndexMetadata` in `@/lib/content/seo`
- * for the shared noindex policy.
- */
 export async function generateMetadata({
   searchParams,
 }: {
   searchParams: Promise<{ page?: string; tag?: string }>
 }): Promise<Metadata> {
   const { page, tag } = await searchParams
-  const pageNum = Math.max(1, Number(page || 1))
+  const posts = await getAllPostMeta()
+  const pageNum = resolveContentPage(posts, { page, tag })
+  if (pageNum === null) notFound()
   return buildIndexMetadata({ tag, pageNum })
 }
 
@@ -25,8 +23,9 @@ export default async function LibraryIndex({
   searchParams: Promise<{ page?: string; tag?: string }>
 }) {
   const { page, tag } = await searchParams
-  const pageNum = Math.max(1, Number(page || 1))
   const posts = await getAllPostMeta()
+  const pageNum = resolveContentPage(posts, { page, tag })
+  if (pageNum === null) notFound()
 
   return (
     <ContentIndexPage

apps/sim/lib/content/index-list.ts

@@ -55,3 +55,14 @@ export function selectVisiblePosts(
   const { featured, remaining } = paginateContentPosts(posts, options)
   return [...featured, ...remaining]
 }
+
+/** Resolves an archive page number, rejecting invalid or empty pagination URLs. */
+export function resolveContentPage(
+  posts: ContentMeta[],
+  { tag, page }: { tag?: string; page?: string }
+): number | null {
+  const pageNum = Number(page || 1)
+  if (!Number.isSafeInteger(pageNum) || pageNum < 1) return null
+  const { totalPages } = paginateContentPosts(posts, { tag, page: pageNum })
+  return pageNum <= totalPages ? pageNum : null
+}

apps/sim/lib/content/seo.ts

@@ -181,11 +181,8 @@ export function buildPostGraphJsonLd(post: ContentMeta, section: ContentSection)
 }
 
 /**
- * Filtered/paginated index variants render genuinely different lists, but
- * only the bare index is indexable — same policy as the integrations and
- * models catalogs — so canonical always points at the unfiltered index and
- * the variant itself is noindexed rather than asking Google to index every
- * tag/page permutation.
+ * Paginated collections contain distinct posts and use their own canonical
+ * URLs. Tag filters remain noindexed to avoid indexing list permutations.
  */
 export function buildIndexMetadata(
   section: ContentSection,
@@ -200,8 +197,8 @@ export function buildIndexMetadata(
     ? `Sim ${section.name.toLowerCase()} posts tagged "${tag}": ${section.description}`
     : section.description
 
-  const canonical = `${SITE_URL}${section.basePath}`
-  const isFiltered = Boolean(tag) || pageNum > 1
+  const isFiltered = Boolean(tag)
+  const canonical = `${SITE_URL}${section.basePath}${!isFiltered && pageNum > 1 ? `?page=${pageNum}` : ''}`
 
   return withFilteredNoindex(
     {
@@ -363,9 +360,8 @@ export function buildAuthorGraphJsonLd(section: ContentSection, author: Author)
  *
  * `tag`/`page` describe which filtered/paginated variant `posts` came from,
  * so `url` reflects the actual page these `posts` are visible on rather than
- * always the bare section index - the same variant is `noindex`ed (see
- * `buildIndexMetadata`), but the graph still shouldn't attribute a partial
- * list to the unfiltered collection URL.
+ * always the bare section index. Only tag-filtered variants are noindexed;
+ * unfiltered pagination uses the same URL in metadata and structured data.
  */
 export function buildCollectionPageJsonLd(
   section: ContentSection,

apps/sim/lib/landing/seo.ts

@@ -89,13 +89,8 @@ export function buildLandingMetadata({
 }
 
 /**
- * Google's documented pattern for faceted/filtered navigation: keep the single
- * unfiltered listing indexable and `noindex` (but still `follow`) any
- * filtered or paginated variant, so link equity flows through without asking
- * Google to index every query-param permutation. Used by every catalog page
- * that serves distinct content per query param (integrations, models, blog,
- * careers, pricing) — `metadata.alternates.canonical` on all of them still
- * points at the bare URL regardless of `isFiltered`.
+ * Noindexes alternate filtered views while keeping their links crawlable.
+ * Callers decide which variants are filters; content pagination is indexable.
  */
 export function withFilteredNoindex(metadata: Metadata, isFiltered: boolean): Metadata {
   return { ...metadata, ...(isFiltered && { robots: { index: false, follow: true } }) }