Scores text based on rule-based patterns to identify formulaic AI writing and provide actionable advice.
Sponsored
Slop Guard is a powerful, rule-based prose linter designed to identify and score text for formulaic AI writing patterns without relying on LLM judges or external API calls. It programmatically runs approximately 80 compiled patterns against your text, providing a numeric score from 0-100. Users receive a detailed list of specific violations with surrounding context and concrete advice for each detected issue, helping improve writing quality and reduce "slop" in various types of content.
Key Features
0152 GitHub stars
02Rule-based prose analysis for formulaic AI writing patterns
03Detailed violation reports with context and actionable advice
04Numeric scoring (0-100) and qualitative bands for text quality
05CLI for linting files and inline text without network calls
06Customizable rule configurations and a tool to fit rules from corpus data
Use Cases
01Lint prose files from the terminal to improve writing quality and consistency
02Implement as a CI gate to enforce writing standards and prevent formulaic text in projects
03Integrate into LLM agents (e.g., Claude, Codex) for real-time writing analysis
