Use case

Stop turning SEO plans into tickets.

You mapped the keywords. You know the pages you need. But implementation still crosses framework details, component reuse, metadata, schema, canonical URLs, internal links, and sitemap updates. Pagewell turns that handoff into an agent workflow with built-in QA.

$ bunx skills add ReScienceLab/pagewell --skill pagewell

Installs the Pagewell agent skill via skills.sh.

View product page

Audience

Built for devs and growth teams

Use Pagewell when marketing pages need to ship faster but still belong in the existing codebase.

Workflow

Keyword to page, automatically

The agent turns intent into a framework-neutral spec, then implements the page using the site adapter and existing components.

Safety

Checked before it goes live

The generated page is checked for metadata, claims, schema, design consistency, genericness, and sitemap behavior.

The bottleneck

Your content map is clear. Implementation is the bottleneck.

A content map may be clear, but implementation still has to cross framework details, component reuse, metadata, schema, canonical URLs, internal links, and discovery files. Pagewell turns that repeated handoff into an agent workflow with explicit gates.

The product truth is not recreated in each prompt; it lives in PAGEWELL.md.
The visual rules are not inferred from memory; they live in DESIGN.md.
The page type is not improvised; it follows a playbook and then adapts to existing page patterns.

The workflow

From brief to reviewed code — in one workflow.

01 brief

State route and intent

The brief names the route, page type, keyword or audience, CTA, noindex decision, and constraints.

02 context

Read PAGEWELL.md and DESIGN.md

The agent gets product claims, prohibited claims, site conventions, route patterns, component patterns, and visual rules.

03 coverage

Map playbook to page architecture

Direct answer, proof, examples, FAQs, and CTAs are coverage jobs, not canned section headings.

04 implementation

Edit the repo

The agent follows Astro, Next.js, Worker, static HTML, or custom conventions already documented in the context.

05 QA

Check the page like a system

Metadata, schema, claim support, internal links, crawlability, private page behavior, and genericness are checked before completion.

06 discovery

Update crawlers when relevant

Indexable routes can be added to sitemap/robots/llms workflows while noindex/private routes stay excluded.

Concrete request

One prompt. A complete page.

example prompt

Use Pagewell to create a use-case page for:
route: /use-cases/generate-seo-pages-as-code
audience: developers and SEO teams
intent: explain how to generate SEO/GEO pages in a repo
CTA: install Pagewell
constraints:
  - use PAGEWELL.md claims only
  - reuse existing Astro components
  - run QA and update discovery if indexable

FAQ

Common questions, answered directly.

Why generate SEO pages as code instead of in a hosted builder?

Code-native generation keeps routes, metadata, components, schema, and discovery files in the same repo your team already reviews and deploys.

How does Pagewell prevent generic SEO pages?

Pagewell uses PAGEWELL.md for product facts and page patterns, treats playbook sections as coverage rather than templates, and adds an anti-generic QA pass.

Can Pagewell work with an Astro site?

Yes. The Astro adapter tells the agent to follow file-based routes, layouts, components, metadata conventions, and static output patterns already present in the repo.

Install Pagewell

Ready to generate SEO pages in your repo?

Install Pagewell and turn page briefs into reviewable routes with metadata, schema, and QA built in.