Skip to content

docs(roadmap): mdx-content-system phase + GSC-validated content paths#500

Draft
bntvllnt wants to merge 2 commits into
mainfrom
worktree-guides-ai-ui-pillar
Draft

docs(roadmap): mdx-content-system phase + GSC-validated content paths#500
bntvllnt wants to merge 2 commits into
mainfrom
worktree-guides-ai-ui-pillar

Conversation

@bntvllnt

@bntvllnt bntvllnt commented Jul 13, 2026

Copy link
Copy Markdown
Collaborator

What

Adds the MDX-max content architecture and the GSC-validated SEO paths to ROADMAP.md, per the roadmap convention. Docs-only — no code, no pages built.

Why

Asked to propose new site paths to improve SEO/impact, grounded in real Search Console data. A 90d GSC pull (sc-domain:ui.vllnt.ai/.com) drove every path below.

New phase — mdx-content-system [PLANNED]

Sequenced before seo-content-engine (which now declares needs: it). 7 tasks:

  • Shared MDX component map (ComponentGrid, CompareTable, Faq, …) — today every caller passes the empty default, which is the real gap
  • Extended pageFrontmatterSchema (faq[], componentSlugs[], type: +guide|vs|react) → JSON-LD emitted by the shell from frontmatter, never the MDX body
  • Generic /guides/[slug], /vs/[slug], /react/[slug] shells → a new page is one .mdx file drop (no TSX, no staticRoutes() edit)
  • Sitemap via content/pages/** glob
  • Author guard for the MDX evaluate() hazard (inline `<X>` code-spans break JSX pages)
  • Migrate the 3 shipped TSX /vs pages onto the MDX shell, URLs unchanged

Infra discovery: MDXContent already supports a components map + enableMDX (auto-detects JSX, falls back to ReactMarkdown+remarkGfm), and getPageContent already falls back <locale>.mdx → default → flat. i18n is opt-in per page: ship en.mdx, add fr.mdx later with zero code change.

seo-content-engine — +7 tasks (.9–.15)

Task Path GSC evidence
.9 /vs/radix ~20 impr, largest uncovered cluster
.10 /vs/mui cross-framework "material design X"
.11 /vs/assistant-ui/alternatives/assistant-ui + 301 namespace split
.12 /react/{callout,floating-action-button,code-block} ~100 impr of react <component>
.13 /guides/ai-ui pillar (+ mobile coming-soon) zero head-term demand — explicit recall/GEO bet
.14 /guides/streaming-ui, /guides/design-tokens evergreen
.15 Validate .9–.14 (E2E)

Namespace decisions recorded

/vs = tools+libs · /alternatives = direct competitors · /integrations = how-to-use · /guides = evergreen reference · /blog = dated articles.

Also

  • Later: multiselect + filter-bar — GSC demand with no component today.

Pre-existing drift found (flagged, NOT fixed here)

  • 2 ACTIVE phases (family-pages, component-sidebar) — convention says exactly one.
  • docs/specs/ doesn't exist — all 5 spec refs dangle; 3 pre-date this PR.

Notes

  • Issue-link CI gate: add Closes #N before merge (kept draft).

90d GSC analysis (sc-domain:ui.vllnt.ai/.com) of real search demand:

- seo-content-engine: SYNC note with the highest-impact CAPTURE paths
  the data proves and that don't exist yet (/alternatives/radix,
  /alternatives/mui, a /react/{component} framework-qualified
  namespace), plus the honest finding that "AI UI development" has
  zero head-term demand today — a recall/GEO bet, not a capture play.
- Later: multiselect + filter-bar component gaps (GSC demand exists,
  no component in the registry today).

Planning/analysis only — no page built (per direction).
@vllnt-pilot
vllnt-pilot Bot deployed to Preview · pr-500 (ntk) July 13, 2026 12:23 Active
@vllnt-pilot

vllnt-pilot Bot commented Jul 13, 2026

Copy link
Copy Markdown

ntk preview

App Status Preview
ui-registry Skipped (preview cell not configured)
storybook Skipped (preview cell not configured)

Built from 04312ea4 · public + no-index (ADR-082) · torn down on close

…aths

Adds the full MDX-max plan and the GSC-validated SEO paths as roadmap
work, per the roadmap convention.

New phase `mdx-content-system` [PLANNED] (7 tasks, sequenced before
seo-content-engine):
- Shared MDX component map, extended frontmatter schema (JSON-LD from
  frontmatter), generic /guides|/vs|/react [slug] shells, sitemap via
  content glob, an author guard for the MDX evaluate() bracket hazard,
  and migration of the 3 shipped TSX /vs pages onto the MDX shell.
- Outcome: a new content page = one .mdx file drop (no TSX, no sitemap
  edit), i18n-ready via fr.mdx with zero code change.

seo-content-engine: +7 tasks (.9-.15) for the GSC-proven paths —
/vs/radix, /vs/mui, /react/{callout,floating-action-button,code-block},
/guides/ai-ui pillar (+ mobile coming-soon), /guides/streaming-ui +
/guides/design-tokens, and the assistant-ui namespace migration.

Namespace decisions recorded: /vs = tools+libs, /alternatives =
direct competitors, /integrations = how-to-use, /guides = evergreen
reference, /blog = dated articles.
@vllnt-pilot
vllnt-pilot Bot deployed to Preview · pr-500 (ntk) July 17, 2026 14:03 Active
@bntvllnt bntvllnt changed the title docs(roadmap): GSC-validated SEO paths + component gaps docs(roadmap): mdx-content-system phase + GSC-validated content paths Jul 17, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant