No AI Slop

No AI Slop is a free, MIT-licensed writing skill created by Peter Yang — a product leader with experience at Roblox, Reddit, Twitch, and Meta — and published on GitHub in July 2026. It installs into AI coding environments like Claude Code or OpenAI Codex and runs as a slash command to detect and remove characteristic patterns of AI-generated writing from any piece of text.

What It Is

No AI Slop is a prompt-based skill file (a SKILL.md with an accompanying eval.md) that you install globally into your preferred AI harness. Once installed, it exposes a /no-ai-slop command that can edit drafts, detect AI writing patterns, or — for satirical purposes — generate maximally clichéd AI slop. It is not a standalone app or SaaS product; it is a reusable instruction set that runs inside an existing AI agent environment.

What It Catches

The skill targets more than 20 specific patterns that commonly appear in AI-generated writing. According to the GitHub README, these include:

• Binary contrasts — "It's not X. It's Y."
• Throat-clearing openers — "Here's the thing..."
• Faux-insight setups — "What nobody tells you..."
• Colon reveals — "The best part: it learns."
• Superficial analysis — "...highlighting the team's commitment to innovation"
• Importance puffery — "marks a pivotal moment," "a testament to"
• Weasel attribution — "experts agree," "studies show"
• Fake-strong verbs — "serves as a centralized hub"
• Synonym cycling — referring to the same thing as "the agent," then "the assistant," then "the tool"
• Dramatic fragmentation — "That's it. That's the whole thing."
• Negative listing — "Not a X. Not a Y. A Z."

Beyond pattern removal, the skill also enforces writing fundamentals: leading with the point, using active voice, untangling complex sentences, and preferring concrete numbers over abstractions.

How It Works

Installation is a single step: paste the GitHub URL into Claude Code, Codex, or another AI harness and tell it to install the skill globally. After that, three usage modes are available:

1. Edit mode — /no-ai-slop [draft] returns the cleaned draft plus a "What changed" summary. The skill is designed to make the minimum effective edit, then self-check its own output against the eval.md pass/fail criteria.
2. Detection mode — /no-ai-slop is this AI slop? [text] lists every pattern found, with the offending quoted line.
3. Satire mode — draft an AI slop post [topic] generates maximally clichéd AI writing for comedic purposes.

Context and Motivation

Peter Yang describes the skill as part of his broader "25/50/25" writing philosophy: the first 25% of writing (rough drafting) and the last 25% (line-by-line manual review) should be done without AI, while AI assists in the middle 50%. The skill is designed to help writers use AI for editing without surrendering voice or quality. Yang notes that Substack shipped its own AI slop detection feature around the same time, and featured his video as an example of responsible AI-assisted editing.

Current Status

The repository was created on July 7, 2026, and last updated on July 23, 2026. According to GitHub metadata, it had accumulated 858 stars and 70 forks within roughly two weeks of launch. The project is actively maintained under the MIT License and is part of Yang's larger "Behind the Craft" personal AI operating system, which includes additional skills, courses, and workflows available through his Substack newsletter.