Pagewell: agent skills for generating pages as code

The problem

Solo founders need more than a homepage. They need product pages, use-case pages, docs, FAQs, glossary entries, free tools, comparison pages, launch pages, and sometimes private outbound pages. Each page needs to match the current site’s framework, design system, product facts, SEO constraints, schema conventions, and discovery files.

Generic AI page generation usually skips those details. It produces isolated copy, ignores the existing codebase, invents claims, or assumes a hosted builder.

What Pagewell does

Pagewell is a framework-agnostic skill pack for coding agents. It captures the workflow behind useful AI page systems without turning your website into a no-code builder.

The Pagewell loop is:

DESIGN.md + PAGEWELL.md + page brief
→ framework-neutral page spec
→ implementation in the current site's stack
→ SEO/GEO/schema/claim QA
→ discovery files updated

The output is regular website code or content in your repository. Pagewell does not require a hosted runtime, visual editor, or proprietary page builder.

What is inside

Pagewell includes a router skill plus core skills for site discovery, design-system extraction, product-fact definition, topic-cluster planning, page generation, free tool generation, QA, discovery-file updates, and SEO operations.

Detailed playbooks cover product pages, use-case pages, FAQ pages, glossary pages, alternative pages, docs, landing pages, private ABM pages, free tools, internal linking, maintenance, and claim checks.

How it fits into a one person company workflow

Use Pagewell when you already have a website repo and want a coding agent to add durable pages without breaking the site’s structure.

A practical workflow looks like this:

1. Consolidate product facts, claim rules, route patterns, and site conventions in PAGEWELL.md.
2. Capture the visual contract in DESIGN.md.
3. Write or infer a page brief for a real search, conversion, support, or outbound intent.
4. Let the agent generate the page, run QA, and update discovery files when the page is indexable.

Best use cases

Pagewell is useful for building topical-authority clusters, publishing product and use-case pages, creating no-signup free tools, keeping docs consistent with product facts, generating campaign landing pages, and building private ABM pages that stay out of search indexes.

It pairs naturally with the rest of the ReScience Lab toolchain: RequestHunt finds demand signals, Pagewell turns validated intents into pages, Hal can implement deeper site changes, NoIdea can supply reusable expertise, and Markshare can publish standalone markdown artifacts.

What makes a good Pagewell brief

A strong brief names the route, search intent, audience, primary CTA, proof constraints, internal links, and any claims the page must avoid. Pagewell is most useful when those inputs are specific enough for an agent to produce a page that fits the site instead of generic marketing copy.

Who it is not for

Pagewell is not for teams that want a hosted no-code editor or one-click publishing system. It is for developers and operators who are comfortable letting an agent work in an existing website repo, then reviewing the generated code, content, metadata, schema, and discovery files.