No AI Slop
An editing skill for Claude Code, Codex and other AI tools that removes more than 20 common AI-writing patterns from a piece of text, such as fake-profound endings and weasel attribution, while keeping the writer's own voice.
No AI Slop is an editing skill for Claude Code, Codex, ChatGPT and other coding or writing agents that removes more than 20 recurring patterns of AI-sounding writing from a piece of text, such as binary contrasts phrased as 'it's not X, it's Y', throat-clearing openers, faux-insight setups, colon reveals, dramatic one-line fragments, importance puffery, weasel attribution and fake-profound endings, while trying to preserve the writer's own vocabulary, cadence and humor rather than flattening it into generic prose.
Used on a piece of writing, the skill can either rewrite it to remove the detected patterns and list what it changed, or only flag which patterns it found without rewriting anything, so a writer can fix the text themselves. It also checks general writing fundamentals, such as leading with the point, using active voice, untangling hard-to-follow sentences and preferring concrete detail over abstraction, and can be asked to generate deliberately exaggerated 'AI slop' text as satire.
It is installed by asking a coding agent to install the skill directly from its GitHub repository, or through a package-based skill installer.