Build Page logo

Build Page

Community
GabrielAmz
build-page

The terminal assembler in Web Anatomy. Take a page structure plus its written copy and assemble them into one shareable wireframe with a coherent, single-exemplar design system (type scale, color direction, spacing, component style). Runs on both branches: a newly created page (build) or an improved existing page. Use when the user asks to assemble the page, build the wireframe, turn this into a page I can share, give me a shareable mockup, apply a design system, or lay this out top to bottom. Reuses the copy and references that write-page already produced; it does not write copy and it does not re-diagnose. Derives the design from ONE exemplar (a benchmark page for a new page, the user's own page when improving), never averaged across companies. Writes a self-contained wireframe under `.webanatomy/build-page/`.

Overview

PublisherGabrielAmz
Repositoryweb-anatomy
Skill namebuild-page
Stars
11
Forks
1
Bundled files
Instructions only
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by GabrielAmz on GitHub. Read the source before you install it.

Installation

Install the Build Page AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/GabrielAmz/web-anatomy/tree/main/skills/build-page
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/GabrielAmz/web-anatomy.git /tmp/web-anatomy
mkdir -p .claude/skills
cp -r /tmp/web-anatomy/skills/build-page .claude/skills/build-page
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Build Page in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Build Page on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Build Page is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Build Page

The last step of the page workflow: turn structure + copy into a wireframe someone can actually look at. write-page produces the words; build-page lays them out in a coherent design and hands the user one shareable artifact.

This is the terminal assembler. It runs on both branches:

  • After a new page is written (create): assemble the chosen sections + their copy into a wireframe.
  • After an existing page is improved: assemble the improved structure + copy into a shareable "here is the better version" wireframe (optional; the user can stop at the copy fixes).

It does not write copy (that is write-page), it does not diagnose (that is audit-page), and it does not invent the section list (that comes from the structure it is handed). It assembles and designs.

The one rule: a single design exemplar

A coherent page cannot be stitched from sections of different companies. Their type, color, and spacing clash, and the result reads as broken. So build-page derives the whole design system from exactly ONE source:

  • Create: one exemplar homepage the user likes, picked via find-examples / search_pages. Use it for the look only.
  • Improve: the user's OWN current page. It already has a coherent design language; keep it and re-skin the improved structure + copy in it.

The exemplar drives DESIGN only: type scale, color direction, spacing rhythm, component style, section layout patterns. Never lift the exemplar's words, claims, or brand name. The copy is the user's, from write-page.

Step 0 - Gather the inputs

build-page needs two things. Do not start without both:

  1. Structure + copy. Prefer the most recent write-page output for this page (.webanatomy/write-page/{page}-*/report-data.json and its recommendations / written sections). If the user is running build-page directly without a prior write-page run, ask them to run write-page first (build or improve mode); this skill assembles, it does not generate the copy.
  2. One design exemplar.
    • Create: ask which exemplar homepage to follow, or surface candidates via find-examples and let the user pick one. Capture its screenshot as the design reference.
    • Improve: use the current page already captured by audit-page / write-page (references/current.png). If none exists, capture it now.

If the user has a brand or design system of their own (tokens, a styleguide, an existing site), prefer it over the exemplar for color and type, and use the exemplar only for layout and section composition. Ask once.

Step 1 - Derive the design system from the exemplar

Read the single exemplar's screenshot and extract a small, coherent system. Keep it to what a wireframe needs:

  • Type: heading scale and body size relationship, weight contrast, and whether it is geometric or humanist (describe, do not guess exact fonts unless visible).
  • Color: background direction (light / dark / tinted), one primary action color, and the contrast level. Direction, not exact hex, unless the exemplar makes it obvious or the user gave tokens.
  • Spacing: section rhythm (tight / generous), container width, and alignment (centered / left).
  • Components: button style, card vs flat, how proof and product visuals are placed.

Write this as a short design-system.md (a handful of lines per dimension). This is the "clear design guidelines" half of the deliverable. Follow the house style for any prose: no em-dashes, never expose benchmark internals.

Do not blend a second source. If a section's copy came from a benchmark example with a different look, that example informed the COPY (via write-page), not the design here.

Step 2 - Assemble the wireframe

Lay the agreed sections out top to bottom, in order, each populated with its write-page copy, skinned in the design system from Step 1.

Output a single self-contained HTML file: inline CSS only, no external dependencies, no CDN links, system fonts (or a close stand-in for the exemplar's type feel). It is a wireframe, a coherent shareable draft, not a pixel-final mockup:

  • Real copy from write-page in every section (never lorem ipsum).
  • Section order and hierarchy from the structure.
  • Layout, spacing, and component style from the design system.
  • Placeholders for assets the user does not have yet (product shots, logos), clearly marked, never faked.
  • An HTML comment above each section noting which exemplar pattern it follows and which write-page reference grounded the copy.

The wireframe is the artifact the user shares. Keep it clean and legible over decorative.

Step 3 - Output

Write:

  • .webanatomy/build-page/{page}-{YYYY-MM-DD}/wireframe.html - the shareable wireframe.
  • .webanatomy/build-page/{page}-{YYYY-MM-DD}/design-system.md - the design guidelines.
  • .webanatomy/build-page/{page}-{YYYY-MM-DD}/references/ - the exemplar screenshot and any section references reused from write-page.

Then hand it over explicitly: say the wireframe is ready, give the full path, and offer to open it (open <path> on macOS, xdg-open on Linux, start on Windows). Propose the next move: tweak a section, swap the exemplar to try a different look, or export the copy and design notes into the user's coding agent to build the real page.

Guardrails

  • One exemplar, always. Never average a design across companies. If the user wants to compare looks, run twice with two exemplars and let them choose.
  • Assemble, do not rewrite. Use write-page's copy verbatim. If the copy needs changing, that is a write-page job, not this one. Do not re-query the benchmark for copy.
  • Design only from the exemplar. Never copy its words, claims, metrics, or brand.
  • No faked proof. Mark missing assets (testimonials, logos, product shots) as placeholders; never invent them.
  • It is a wireframe. Coherent and shareable, not pixel-final. Say so when handing it over.
  • House style. Follow webanatomy-setup/references/house-style.md for any prose: no em-dashes, "The X…" not "Your X…", and never expose benchmark scores, thresholds, or internal fields. Answer in the user's language.

Future enhancement

V1 writes the HTML wireframe directly. A deterministic wireframe renderer (a script that takes structure + copy + design tokens and emits the HTML, mirroring the report renderer the other skills use) is a later improvement; until then, keep the hand-written HTML clean and self-contained.

Frequently asked questions

What does the Build Page AI skill do?

The terminal assembler in Web Anatomy. Take a page structure plus its written copy and assemble them into one shareable wireframe with a coherent, single-exemplar design system (type scale, color direction, spacing, component style). Runs on both branches: a newly created page (build) or an improved existing page. Use when the user asks to assemble the page, build the wireframe, turn this into a page I can share, give me a shareable mockup, apply a design system, or lay this out top to bottom. Reuses the copy and references that write-page already produced; it does not write copy and it doe...

Why use Build Page on TypingMind?

Because you install it once and use it with any model. Build Page is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Build Page in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/GabrielAmz/web-anatomy/tree/main/skills/build-page. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Build Page?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Build Page?

As many as you like. As long as a model supports skills, you can use Build Page with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Build Page AI skill free?

Yes. It is published on GitHub by GabrielAmz under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇