docwriter-org/plain-writing-skill

A plain-language writing skill for AI agents, with a revision view that shows what changed.

What it solves

It prevents AI agents from producing "slop"—text that is overly complex, assumes too much context, or uses unexplained technical terms—by enforcing a plain writing style.

How it works

The project provides a set of plain-text rules and steps (contained in a markdown file) that can be integrated into an AI agent's system prompt or rules file. The agent follows these guidelines to write and revise text, and it can specifically be triggered via a /plain-writing deslopify command to rewrite previous responses for clarity and conciseness.

Who it’s for

Users of AI agents (such as Claude Code, Codex, or pi) who want their AI-generated content to be clear, professional, and accessible to readers without deep project context, such as executives.

Highlights

  • Agent-agnostic: Works with any agent that can read a markdown file or accept system instructions.
  • Self-correcting: The skill includes instructions for the agent to check its own writing and remove unnecessary content.
  • "Deslopify" command: A dedicated command to rewrite unclear or overly technical responses into a clear structure.
  • Evaluation suite: Includes optional evaluation tools to measure the effectiveness of the writing skill.

Related

  • Project
  • Project
  • Project
  • Project
  • Project