playbooks

home / skills / cdeistopened / opened-vault / human-writing-copy

human-writing-copy skill

/.claude/skills/_archived/human-writing-copy

This skill helps transform AI-assisted drafts into authentic, distinctive prose by rewriting from core message and preserving voice.

npx playbooks add skill cdeistopened/opened-vault --skill human-writing-copy

Review the files below or copy the command above to add this skill to your agents.

Files (5)
SKILL.md
2.9 KB
---
name: human-writing
description: Transform AI-assisted drafts into authentic, distinctive prose. Use this skill when (1) reviewing or humanizing AI-generated content, (2) writing content that must not appear AI-assisted, (3) converting transcripts or source material into polished prose, (4) maintaining Charlie Deist's voice across content. Triggers on requests to "make it sound human", "remove AI tells", "write in my voice", or any content editing/writing task.
---

# Human Writing

Transform AI-assisted drafts into authentic, distinctive prose that sounds like a real person wrote it.

## Core Principle

**Rewrite, don't patch.** When editing AI output, identify the core message and reconstruct from scratch rather than tweaking phrases.

## The #1 AI Tell

**The Correlative Construction:** "X isn't just Y - it's Z"

This is the default structure AI produces. Every AI sample uses it. Eliminate it completely.

- Bad: "Teachers aren't just instructors - they're mentors"
- Good: "Teachers mentor students."

## Forbidden Patterns

### Constructions
- "It's not about X, it's about Y"
- "The truth is..." / "The reality is..."
- "Now more than ever..."

### Rhetorical Flourishes
- "The best part? ..."
- "Here's the thing..."
- "Let's be honest..."
- "What if I told you..."

### Faux-Dramatic Fragments
- "No fluff. No filler. Just results."
- "Simple. Clear. Effective."

### Openers
- "In the ever-evolving world of..."
- "In today's fast-paced..."
- "Gone are the days when..."

### Formatting
- No boldface for emphasis in body text
- No emojis unless explicitly requested
- No "In conclusion" or "In summary"
- Use hyphens with spaces - like this - not em dashes

## Quick Scan Checklist

Before publishing, check for:
- [ ] Correlative constructions ("not just X, but Y")
- [ ] "Just" appearing more than once
- [ ] Hedge words (might, could, perhaps, possibly)
- [ ] Passive voice
- [ ] Vague claims without specifics
- [ ] Forbidden rhetorical flourishes

## Charlie's Signature Moves

1. **Aphoristic Opening** - Thesis as proverb
2. **Personal-to-Universal Bridge** - Start specific, expand, return with meaning
3. **Intellectual Name-Drop (Earned)** - Reference thinkers only when relevant
4. **Contrarian Hook** - Challenge received wisdom without edgelording
5. **Incarnational Detail** - Specific, sensory, grounding
6. **Self-Deprecating Confession** - Prevents preachiness, builds trust

## Target Tone

A coach who has done the thing and believes you can too. A friend who tells you the truth because he respects you. A fellow traveler, slightly ahead on the path.

## Extended References

For complete documentation, see:
- `references/forbidden-patterns.md` - Full list of AI tells
- `references/sticky-sentences.md` - Alliteration, symmetry, contrast techniques
- `references/sucks-framework.md` - Pre-writing checklist
- `references/humanization-workflow.md` - Step-by-step editing process

Overview

This skill transforms AI-assisted drafts into authentic, distinctive prose that reads like a real person wrote it. I rebuild unclear or mechanical AI output into a coherent voice while preserving the original intent. It specializes in preserving a specific signature voice across formats and can convert transcripts or rough drafts into polished copy.

How this skill works

I scan the draft for AI tells, weak constructions, passive voice, and vague claims, then rewrite from the core message instead of patching phrases. I apply voice-guided techniques—specific sensory detail, contrarian hooks, earned references, and brief self-aware confessions—to produce readable, human-centered prose. The process includes a quick checklist pass and a final read for cadence and clarity.

When to use it

  • Humanize AI-generated articles, emails, or landing page copy
  • Convert interview transcripts or notes into clean, publishable prose
  • Produce content that must not read like it was AI-assisted
  • Maintain a consistent authorial voice across posts or formats
  • Tighten messaging for public-facing content that needs credibility

Best practices

  • Rewrite around the core idea instead of editing isolated sentences
  • Remove correlative constructions like "not just X - but Y" and other repeatable AI tells
  • Favor active voice and concrete specifics over hedging language
  • Use short, varied sentence lengths to create natural cadence
  • Run the quick scan checklist before finalizing: passive voice, hedges, vague claims, repeated "just"

Example use cases

  • Turn a verbose AI outline into a crisp blog post in Charlie Deist's voice
  • Edit marketing copy so it sounds conversational and earned for a product launch
  • Refine an internal memo or policy summary to read like a colleague wrote it
  • Convert a recorded interview into a feature-style article with sensory detail
  • Polish social posts and newsletters to remove mechanical phrasing and increase engagement

FAQ

Will you preserve the original meaning when rewriting?

Yes. I identify the draft's core message first and reconstruct the prose around it so meaning and intent remain intact.

Can you match a specific writer's voice consistently?

Yes. I use signature moves—specific detail, contrarian hooks, and measured self-disclosure—to maintain a repeatable voice across pieces.