avoid-ai-writing

Detect and rewrite AI writing patterns in prose using severity tiers and vocabulary tables.

Updated Jul 28, 2026
One-click install
npx skills add https://github.com/christian-byrne/comfy-skills --skill avoid-ai-writing-christian-byrne
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: avoid-ai-writing
Source: https://github.com/christian-byrne/comfy-skills/tree/main/skills/avoid-ai-writing
Command: npx skills add https://github.com/christian-byrne/comfy-skills --skill avoid-ai-writing-christian-byrne

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? AI-generated text carries recognizable patterns—chatbot artifacts, inflated vocabulary like "delve" and "leverage," em dash overuse, and formulaic structures—that undermine credibility with readers. This Skill audits content for 40 documented AI-isms and rewrites it to sound human. ## Core Features & Use Cases - Two modes: rewrite (default) flags and fixes AI patterns; detect flags only, for audits where the writer decides what to change. - Severity-tiered rules: P0 credibility killers (cutoff disclaimers, chatbot artifacts), P1 obvious AI smell (word-list violations, template phrases), and P2 stylistic polish, with a three-tier vocabulary replacement table. - Context profiles: Auto-detects content type (LinkedIn post, blog, technical blog, investor email, docs) and relaxes or tightens rules per a tolerance matrix. - Use Case: Paste a draft blog post and ask for an audit—the Skill quotes every AI-ism found, produces a rewritten version, summarizes the edits, and runs a second-pass audit to catch surviving patterns. ## Quick Start Audit this blog post draft for AI writing patterns and rewrite it to sound human.

Frequently Asked Questions about avoid-ai-writing

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

FAQPage Schema
How do I remove AI writing patterns from my text?▼

Paste your text and request a rewrite. The Skill flags every AI-ism with the offending text quoted, produces a clean rewritten version preserving your structure and details, summarizes the changes, and runs a second-pass audit to catch surviving patterns.

How to detect AI-generated text without rewriting it?▼

Use detect mode by saying "detect," "flag only," or "audit only." The Skill groups every flagged AI-ism by severity (P0, P1, P2) and gives an assessment of which flags are clear problems versus judgment calls, without altering your text.

What words and phrases signal AI-generated writing?▼

High-signal words include delve, leverage, robust, tapestry, realm, seamless, and testament to, which appear 5-20x more often in AI text. Structural signals include em dash overuse, "It's not X—it's Y" constructions, rule-of-three groupings, and uniform sentence length.

Does the AI writing audit work differently for LinkedIn posts versus blogs?▼

Yes. The Skill auto-detects context (LinkedIn, blog, technical blog, investor email, docs) and applies a tolerance matrix. LinkedIn posts get relaxed em dash, bold, and bullet rules, while investor emails get extra-strict promotional language and significance inflation checks.

When should text be fully rewritten instead of patched for AI patterns?▼

A full rewrite is advised when text has all three: 5+ flagged vocabulary hits across categories, 3+ distinct pattern categories triggered, and uniform sentence and paragraph length. At that point the structure itself is AI-generated and patching will not fix it.