Introduction

Web Anatomy gives your AI coding agent live access to a benchmark library of real, high-performing websites, so it recommends, writes, and edits from proven examples instead of guessing.

It ships in two parts:

• Skills — named workflows your agent runs (audit, improve, compare, find examples). Free, open, and installed in one command. No account, no token.
• The MCP server — the live link to the benchmark library, included in the Pro plan, so your agent can search scored sections, fetch real screenshots, and benchmark a live URL on demand.

The skills work on their own. The MCP adds the live data.

Why use it

• Grounded in real pages. Every suggestion points to a real, high-performing website your agent can show you.
• Skip the blank page. Find proven hero, pricing, CTA, and FAQ patterns by industry in seconds.
• Inside the agent you already use. It works in Claude Code, Cursor, Codex, and any agent that speaks MCP.
• Outputs your team can ship. Turn the findings into a clean report or a prioritized fix list.

What your agent can do

• Find proven section examples filtered by industry and locale.
• Improve a page section against stronger references before editing copy or layout.
• Benchmark a live URL or screenshot and get prioritized fixes.
• Turn references into a short report your team can act on.