anti-ai-slop
github.com/misbahsy/anti-ai-slop- Category
- AI Agents
- Pricing
- Open Source
- Type
- TOOL
- GitHub
- 4 stars
- Date
About
anti-ai-slop is a skill for Claude Code, Codex, and other coding agents that detects and removes AI writing tells from a draft, checking for roughly 120 flagged words and 100 phrases along with patterns like em dashes, colon reveals, throat-clearing openers, and negative parallelism. It runs text through four gates, including an agent review step that compares the original against the rewrite and loops back on failed reviews, plus an optional second-model review for scoring.
Why it made the leaderboard
Gives coding agents a repeatable, multi-pass check to strip AI writing tells like em dashes and colon reveals from generated prose before it ships, rather than relying on a single style-prompt.
Tags
Comments (0)
No comments yet
Editorially curated, with community endorsements as a secondary signal. Corrections welcome.