This one isn’t mine. It was made by Peter Yang, who shares it free, and I use it on almost everything I write with AI, including this website. All the credit is his, so get it from him: his GitHub page, or read his write-up of how it works.
What it does
You’ve seen AI writing even if you can’t say why. Everything is “not just X, it’s Y”. Every paragraph ends on a neat little punchline. Words like “delve”, “leverage” and “game-changing” turn up everywhere. People notice, and they stop trusting what they’re reading.
No AI Slop edits your draft the way a sharp human editor would.
- Takes out the AI patterns, more than twenty of them, from the tired phrases to the fake-dramatic sentence shapes.
- Keeps your voice. It makes the smallest change that fixes the problem, so the result still sounds like you, not like polished corporate copy.
- Tells you what it changed, in a short list after the edited draft.
- Can just check instead of rewriting. Ask whether something reads like AI and it points to each line that gives it away, with a quick fix for each.
When to use it
Emails, social posts, website copy, proposals, newsletters. Anything where people need to believe a person wrote it.
How to get it
Peter shares it on GitHub. There are two ways to install it.
If you use Claude Code, start a chat and paste this in, word for word. Claude does the rest.
Install the /no-ai-slop skill globally from https://github.com/petergyang/no-ai-slop
If you use Claude in the browser or the Claude app, go to Peter’s GitHub page for the skill files. Skills are installed in Claude under Settings, then Capabilities, then Skills, the same way as the other skills here.
Once it’s in, paste a draft into Claude and ask: “Run no AI slop over this.”