Agents and developers

How AI agents read the Womp docs, use the Womp skill, and, soon, act in the Womp editor.

Use this when: you connect an AI agent to Womp, or you are an agent deciding where to read.

Don't use this for: using Womp yourself. Start at Get started.

The Womp docs are written for people and agents. Every page is Markdown, every design-rule number has a rule ID, and the rules are published as data.

What is available

The top bar of a docs page with buttons for Copy page, Markdown, ChatGPT and Claude.

Every docs page has Copy page, a Markdown button, ChatGPT and Claude in its top bar (ChatGPT and Claude are hidden on phones). The AI buttons open the assistant with the page's .md file.

Tool	What it does	Page
Markdown pages	Drop the trailing slash and add `.md` to any docs path. `Copy page`, `ChatGPT` and `Claude` buttons are in each page's top bar	llms.txt and Markdown
`/llms.txt` and `/llms-full.txt`	An index by task, and every page in one file	llms.txt and Markdown
`/docs/data/print-rules.json`	Every published print limit with its rule ID	Design rules
Womp skill	A `SKILL.md` for Claude, Cursor and ChatGPT Skills	Womp skill
WebMCP	Will let a browser agent act in the Womp editor. Not open to all accounts yet	WebMCP

Rules for agents

Read the page for the task before you answer.
Quote print limits with their rule ID, for example CLR-04.
When a page says "being confirmed", say so. Do not guess the value.

Use instead

Wrong move: scraping the HTML pages. Why: the HTML adds navigation and layout. Right move: read the .md twin or /llms-full.txt.

Last checked: 2026-11-09. Womp version: Womp web app, November 2026.