Humanizer enhanced: remove AI writing patterns

Identify and remove signs of AI-generated text. This enhanced version includes crypto/Web3 patterns and adds personality to make content sound genuinely human-written.

Quick start

/humanizer                    # Humanize current file or selection
/humanizer path/to/file.md    # Humanize specific file
/humanizer --scan             # Scan only, don't edit (show issues)
/humanizer --batch drafts/    # Process all .md files in directory

---

Process

Step 1: Scan for patterns

Identify all AI patterns in the text, categorize by severity:

• HIGH — Obvious AI tells, must fix (negative parallelism, chatbot artifacts, em dash overuse, vague attributions, copula avoidance)
• MEDIUM — Common AI patterns, should fix (rule of three, significance inflation, synonym cycling)
• LOW — Minor tells, fix if time permits (title case headings, excessive bold)

Step 2: Report findings

Show user a summary:

## Humanizer scan results

HIGH (3 issues)
- Line 45: Negative parallelism "isn't X. It's Y"
- Line 89: Em dash overuse (5 instances)
- Line 120: "Research shows" without attribution

MEDIUM (5 issues)
- Line 23: Rule of three pattern
- Line 67: Copula avoidance "serves as"
...

LOW (2 issues)
- Line 12: Title case heading
...

Total: 10 issues found
Estimated humanization: ~15 edits needed

Step 3: Fix (with user approval)

Ask user: "Fix all issues? Or review one by one?"

Step 4: Add soul

After fixing patterns, review for personality. Sterile writing is still obvious AI. See references/communication-crypto-soul-patterns.md for the full soul/personality guide.

Step 5: Readability check

Check Flesch-Kincaid readability. Target grade 10-12 for developer content, grade 8-10 for general audience. If score is too high (too complex), simplify longest sentences and replace jargon.

Step 6: Em dash regression scan

After all other fixes, run a final check for em dashes (—) across the text. Humanizer rewrites can reintroduce em dashes. Remove any that were added during the fix process.

---

Pattern routing table

All 34 patterns are documented with before/after examples in the reference files below.

Patterns	Severity	Reference file
1. Significance inflation	MEDIUM	references/content-patterns.md
2. Promotional language	MEDIUM	references/content-patterns.md
3. Superficial -ing analyses	MEDIUM	references/content-patterns.md
4. Vague attributions	HIGH	references/content-patterns.md
5. Formulaic challenges sections	MEDIUM	references/content-patterns.md
6. Generic positive conclusions	MEDIUM	references/content-patterns.md
7. AI vocabulary words	MEDIUM	references/language-style-patterns.md
8. Copula avoidance	HIGH	references/language-style-patterns.md
9. Negative parallelism	HIGH	references/language-style-patterns.md
10. Rule of three	MEDIUM	references/language-style-patterns.md
11. Synonym cycling	MEDIUM	references/language-style-patterns.md
12. False ranges	LOW	references/language-style-patterns.md
13. Em dash overuse	HIGH	references/language-style-patterns.md
14. Excessive boldface	LOW	references/language-style-patterns.md
15. Inline-header lists	MEDIUM	references/language-style-patterns.md
16. Title case headings	LOW	references/language-style-patterns.md
17. Curly quotes	LOW	references/language-style-patterns.md
18. Chatbot artifacts	HIGH	references/communication-crypto-soul-patterns.md
19. Knowledge cutoff disclaimers	HIGH	references/communication-crypto-soul-patterns.md
20. Sycophantic tone	MEDIUM	references/communication-crypto-soul-patterns.md
21. Excessive hedging	MEDIUM	references/communication-crypto-soul-patterns.md
22. Filler phrases	MEDIUM	references/communication-crypto-soul-patterns.md
23. Crypto hype language	HIGH	references/communication-crypto-soul-patterns.md
24. Vague "ecosystem" claims	MEDIUM	references/communication-crypto-soul-patterns.md
25. Unsubstantiated stats	HIGH	references/communication-crypto-soul-patterns.md
26. "Seamless" and "frictionless"	MEDIUM	references/communication-crypto-soul-patterns.md
27. Abstract "empowerment" language	MEDIUM	references/communication-crypto-soul-patterns.md
28. Fake decentralization claims	HIGH	references/communication-crypto-soul-patterns.md
29. Meta-narration	HIGH	references/communication-crypto-soul-patterns.md
30. False audience range	MEDIUM	references/communication-crypto-soul-patterns.md
31. Parenthetical definitions	MEDIUM	references/communication-crypto-soul-patterns.md
32. Sequential numbering	MEDIUM	references/communication-crypto-soul-patterns.md
33. "It's worth noting" filler	MEDIUM	references/communication-crypto-soul-patterns.md
34. Identical paragraph structure	HIGH	references/communication-crypto-soul-patterns.md
Soul and personality guide	—	references/communication-crypto-soul-patterns.md

---

Severity reference

Severity	Patterns	Action
HIGH	Negative parallelism, em dash overuse, chatbot artifacts, vague attributions, copula avoidance, crypto hype, unsubstantiated stats, meta-narration, identical paragraph structure, fake decentralization, knowledge cutoff disclaimers	Must fix
MEDIUM	Rule of three, significance inflation, promotional language, -ing analyses, AI vocabulary, sycophantic tone, hedging, filler phrases, ecosystem claims, false audience range, parenthetical definitions, sequential numbering, "it's worth noting" filler, inline-header lists, "seamless"/"frictionless", abstract empowerment	Should fix
LOW	Title case, curly quotes, excessive bold, false ranges	Fix if time permits

---

Quick reference: find and replace

Find	Replace
— (em dash, multiple)	, or .
serves as / stands as	is
isn't X. It's Y	Rewrite as single statement
crucial / vital / pivotal	important or key or delete
Furthermore, / Moreover,	Also, or delete
It is important to note	Delete
Research shows	Add specific source
landscape (abstract)	Be specific
revolutionizing / game-changing	Describe what it actually does
seamless / frictionless	Describe the actual UX
In this article, we'll explore	Delete
Let's dive in / Let's take a look	Delete
First,... Second,... Third,...	Vary transitions
It's worth noting / Notably,	Delete
delve	"look at" / "examine"
Additionally	Delete

---

Batch mode

To humanize multiple files:

# Scan all markdown files in drafts/
/humanizer --scan drafts/*.md

# Fix all files (with confirmation)
/humanizer --batch drafts/

Output format for batch:

## Batch humanization report

drafts/post-1.md
   HIGH 3 | MEDIUM 5 | LOW 2

drafts/post-2.md
   HIGH 1 | MEDIUM 3 | LOW 4

drafts/post-3.md
   Clean! No issues found.

Total: 3 files, 18 issues

---

Sources

Based on:

• Wikipedia: Signs of AI writing
• GitHub: blader/humanizer
• Original research on crypto/Web3 AI patterns

Key insight: "LLMs use statistical algorithms to guess what should come next. The result tends toward the most statistically likely result that applies to the widest variety of cases."

---

Version 1.2.0 | Created for 0G Labs content team