anti-ai-writing

Rewrite user-facing prose for clarity and accountability across agent surfaces.

Updated Mar 18, 2026
One-click install
npx skills add https://github.com/Canepro/codex-skills --skill anti-ai-writing-canepro
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: anti-ai-writing
Source: https://github.com/Canepro/codex-skills/tree/main/skills/anti-ai-writing
Command: npx skills add https://github.com/Canepro/codex-skills --skill anti-ai-writing-canepro

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

ALWAYS use as the final pass before returning user-visible prose in any agent session (Cursor, Codex, Claude). Direct, human, specific copy for docs, README, site content, commits, PRs, comments, handoffs, and chat replies. Supports rewrite, detect-only, and edit-in-place modes.

Core Features & Use Cases

  • Finishing pass that trims filler and sharpen claims.
  • Rewrites for clarity, conciseness, and accountability.
  • Detect-only mode to audit prose without changes.
  • Edit-in-place mode to patch in-file content.

Quick Start

Run this skill as the final pass on your draft to produce direct, human, and clear prose.

Frequently Asked Questions about anti-ai-writing

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I make AI-generated chat responses sound more human and direct?

You can finalize human-centered prose by running a copy-editing pass on your draft. This finishing step trims filler, sharpens claims, and enforces voice consistency, preventing the removal of essential nuance across docs, READMEs, and chat replies.

How do I finalize prose in READMEs and PR bodies before committing?

You finalize prose in READMEs and PR bodies by applying an edit-in-place finishing pass. This step patches in-file content directly, trimming filler and sharpening claims to ensure your documentation is clear and write-ready.

Can I audit documentation for AI writing without modifying the original text?

Yes, you can audit documentation for AI writing without modifying the original text by using a detect-only mode. This mode evaluates user-facing content to identify filler and voice inconsistencies without applying any edits.

Does the copy-editing finishing pass work across agent surfaces like Cursor and Codex?

The copy-editing finishing pass works across agent surfaces like Cursor, Codex, and Claude. It serves as a final pass to process user-visible prose in chat responses, handoffs, and site content across these environments.

What is the best way to rewrite documentation for clarity and accountability?

The best way to rewrite documentation for clarity and accountability is to apply a human-centered finishing pass. This approach sharpens claims, trims filler, and enforces voice consistency without removing essential nuance or user intent.