Humanizer
Writing & Prompts
Rewrites AI text so it reads like a person wrote it, without changing what it says.
Strips predictable AI phrases, clichéd structures and em dashes from Claude's writing.
One of Sam's favourites. AI writing has patterns: throat-clearing openers, "it's not just X, it's Y", rhetorical questions, dramatic one-line fragments and, yes, em dashes everywhere. Stop Slop teaches Claude to catch and remove them.
It ships three reference files: banned phrases, structural clichés and before-and-after examples. Claude then scores each draft out of 50 on directness, rhythm, trust, authenticity and density, and revises anything that scores under 35.
It is small and simple, which is why it works in Claude Code, Claude Projects and plain API calls alike.
Clone github.com/hardikpandya/stop-slop and copy the folder into ~/.claude/skills/stop-slop so Claude Code picks it up. In claude.ai, upload SKILL.md and the references folder to a Project's knowledge, or paste the core rules into your custom instructions.
Use stop-slop to rewrite this Instagram caption so it sounds like a person: [paste caption]Anyone whose readers can tell when a post was written by AI.
Why it made the list
It is the quickest fix for the em-dash problem and it makes everything else Claude writes sharper too.
Writing & Prompts
Rewrites AI text so it reads like a person wrote it, without changing what it says.
Writing & Prompts
Turns a rough request into an expert prompt for Claude, ChatGPT, Midjourney and 20+ tools.
Productivity
Forces Claude to lead with the answer, number the steps and skip the padding.