Easy Read Skill for Claude

Make Easy Read documents inside Claude. Give Claude some text, a PDF or a Word file. You get back a Word document with short sentences and a picture beside every sentence.

Download the skill (free) How to install it

A .zip file of about 5 MB. Always the latest version — see what changed.

What the skill does

The skill uses the same Easy Read rules and picture library as this website. When you ask Claude to make something Easy Read, it:

  1. Reads your text, PDF or Word file
  2. Rewrites it in short sentences, with one idea in each sentence
  3. Picks a picture for every sentence from over 370 pictures
  4. Checks the result against the Easy Read rules and fixes any problems
  5. Gives you a Word document to download, edit and print

Website or Claude skill?

Both are free. Use whichever suits you.

This website Claude skill
Account needed No Yes, a Claude account
What you give it Pasted text, up to 30,000 characters Pasted text, PDF or Word files
What you get A web page to copy, print or share for 24 hours A Word document
Changes afterwards Edit the text yourself Ask Claude, for example "make the second section shorter"

What you need

How to install it

  1. Download the skill. Do not unzip it.
  2. In Claude, go to Customize → Skills.
  3. Select +, then Create skill, then Upload a skill.
  4. Choose the easy-read.zip file you downloaded.
  5. Check that the skill is turned on in your list of skills.

Try it

Start a new chat in Claude and try one of these:

Your privacy

When you use the skill, your text stays in your own Claude account. It is not sent to this website or to OpenAI. Anthropic, the company that makes Claude, handles it under your Claude account's terms and privacy settings.

Licence and picture credits

The skill is free and open source under the MIT License. You can share it with anyone.

The pictures come from Mulberry Symbols, OpenMoji and the Easy Read Online Dictionary, under Creative Commons share-alike licences. Every document the skill makes ends with a line crediting the pictures it uses. Please keep that line.

The code is on GitHub. Questions or feedback? Email me.