# Send Operator: Humanizer to your agent
Hand the extracted package to your coding agent with a concrete install brief instead of figuring it out manually.
## Fast path
- Download the package from Yavira.
- Extract it into a folder your agent can access.
- Paste one of the prompts below and point your agent at the extracted folder.
## Suggested prompts
### New install

```text
I downloaded a skill package from Yavira. Read SKILL.md from the extracted folder and install it by following the included instructions. Tell me what you changed and call out any manual steps you could not complete.
```
### Upgrade existing

```text
I downloaded an updated skill package from Yavira. Read SKILL.md from the extracted folder, compare it with my current installation, and upgrade it while preserving any custom configuration unless the package docs explicitly say otherwise. Summarize what changed and any follow-up checks I should run.
```
## Machine-readable fields
```json
{
  "schemaVersion": "1.0",
  "item": {
    "slug": "operator-humanizer",
    "name": "Operator: Humanizer",
    "source": "tencent",
    "type": "skill",
    "category": "AI 智能",
    "sourceUrl": "https://clawhub.ai/Kevjade/operator-humanizer",
    "canonicalUrl": "https://clawhub.ai/Kevjade/operator-humanizer",
    "targetPlatform": "OpenClaw"
  },
  "install": {
    "downloadUrl": "/downloads/operator-humanizer",
    "sourceDownloadUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=operator-humanizer",
    "sourcePlatform": "tencent",
    "targetPlatform": "OpenClaw",
    "packageFormat": "ZIP package",
    "primaryDoc": "SKILL.md",
    "includedAssets": [
      "SKILL.md",
      "references/examples.md",
      "references/patterns.md",
      "references/personality-injection.md",
      "references/statistical-signals.md",
      "references/vocabulary.md"
    ],
    "downloadMode": "redirect",
    "sourceHealth": {
      "source": "tencent",
      "slug": "operator-humanizer",
      "status": "healthy",
      "reason": "direct_download_ok",
      "recommendedAction": "download",
      "checkedAt": "2026-05-03T00:45:11.942Z",
      "expiresAt": "2026-05-10T00:45:11.942Z",
      "httpStatus": 200,
      "finalUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=operator-humanizer",
      "contentType": "application/zip",
      "probeMethod": "head",
      "details": {
        "probeUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=operator-humanizer",
        "contentDisposition": "attachment; filename=\"operator-humanizer-2.0.0.zip\"",
        "redirectLocation": null,
        "bodySnippet": null,
        "slug": "operator-humanizer"
      },
      "scope": "item",
      "summary": "Item download looks usable.",
      "detail": "Yavira can redirect you to the upstream package for this item.",
      "primaryActionLabel": "Download for OpenClaw",
      "primaryActionHref": "/downloads/operator-humanizer"
    },
    "validation": {
      "installChecklist": [
        "Use the Yavira download entry.",
        "Review SKILL.md after the package is downloaded.",
        "Confirm the extracted package contains the expected setup assets."
      ],
      "postInstallChecks": [
        "Confirm the extracted package includes the expected docs or setup files.",
        "Validate the skill or prompts are available in your target agent workspace.",
        "Capture any manual follow-up steps the agent could not complete."
      ]
    }
  },
  "links": {
    "detailUrl": "https://openagent3.xyz/skills/operator-humanizer",
    "downloadUrl": "https://openagent3.xyz/downloads/operator-humanizer",
    "agentUrl": "https://openagent3.xyz/skills/operator-humanizer/agent",
    "manifestUrl": "https://openagent3.xyz/skills/operator-humanizer/agent.json",
    "briefUrl": "https://openagent3.xyz/skills/operator-humanizer/agent.md"
  }
}
```
## Documentation

### Humanizer Pro

Transform AI-generated text into authentic human writing by removing robotic patterns and injecting natural personality.

### What This Skill Does

Core Capabilities:

Detects 24 AI writing patterns (significance inflation, promotional language, filler phrases, etc.)
Identifies 500+ AI vocabulary terms across 3 severity tiers
Analyzes statistical signals (burstiness, type-token ratio, sentence uniformity)
Removes chatbot artifacts and sycophantic tone
NEW: Injects personality through parenthetical asides, strategic typos, tangents, and random thoughts

Based on: Wikipedia's "Signs of AI writing", Copyleaks research, and real-world pattern analysis.

### Basic Humanization

When asked to humanize text:

Scan for patterns → Check all 24 patterns in references/patterns.md
Check vocabulary → Flag Tier 1/2/3 terms from references/vocabulary.md
Analyze statistics → Compute burstiness, TTR, sentence variance
Rewrite → Remove AI-isms, add personality
Verify → Read aloud, check if it sounds natural

### Adding Personality

Use the new personality injection features from references/personality-injection.md:

Parenthetical asides → (honestly, this part gets me every time)
Strategic misspellings → Natural typos that don't hurt credibility
Tangents → "wait, that reminds me..." moments
Random thoughts → Going off script with authentic reactions

### Always Check (Core Patterns)

Tier 1 vocabulary → DEAD GIVEAWAYS, ban completely (delve, tapestry, vibrant, seamless, etc.)
Filler phrases → "in order to" → "to", "due to the fact that" → "because"
Chatbot artifacts → "Great question!", "I hope this helps!", "Let me know if..."
Generic conclusions → "The future looks bright", "Exciting times lie ahead"

### Check When Relevant

Promotional language → If writing about places, products, services
Significance inflation → If discussing history, events, milestones
Vague attributions → If making claims without sources
Em dash overuse → If text has many — dashes — everywhere

### Advanced Analysis

For comprehensive humanization or when scoring text:

Read references/patterns.md → All 24 patterns with examples
Read references/vocabulary.md → Complete AI term database
Read references/statistical-signals.md → Burstiness, TTR, readability formulas
Read references/personality-injection.md → How to add human touches

### Write Like a Human, Not a Press Release

Use "is" and "has" freely — "serves as" is pretentious
One qualifier per claim — don't stack hedges
Name your sources or drop the claim
End with something specific, not vague optimism

### Add Actual Personality

Have opinions → React to facts, don't just report them
Vary rhythm → Short. Then longer ones that meander a bit.
Acknowledge complexity → "I genuinely don't know how to feel about this"
Let mess in → Perfect structure feels algorithmic
Use contractions → "don't", "won't", "it's" (natural speech)
Sentence fragments → When it makes sense. Like this.

### Cut the Fat

Remove these automatically:

"In order to" → "to"
"Due to the fact that" → "because"
"At this point in time" → "now"
"It is important to note that" → (just say it)
"In the event that" → "if"

### The 24 Patterns (Quick Reference)

For full details with examples, see references/patterns.md.

#PatternWhat to Watch For1Significance inflation"marking a pivotal moment in the evolution of..."2Notability name-droppingListing media outlets without specific claims3Superficial -ing analyses"...showcasing... reflecting... highlighting..."4Promotional language"nestled", "breathtaking", "stunning", "renowned"5Vague attributions"Experts believe", "Studies show", "Industry reports"6Formulaic challenges"Despite challenges... continues to thrive"7AI vocabulary (500+ words)"delve", "tapestry", "landscape", "showcase"8Copula avoidance"serves as", "boasts" instead of "is", "has"9Negative parallelisms"It's not just X, it's Y"10Rule of three"innovation, inspiration, and insights"11Synonym cycling"protagonist... main character... central figure..."12False ranges"from the Big Bang to dark matter"13Em dash overuseToo many — dashes — everywhere14Boldface overuseMechanical emphasis everywhere15Inline-header lists"- Topic: Topic is discussed here"16Title Case headingsEvery Main Word Capitalized17Emoji overuse🚀💡✅ decorating professional text18Curly quotes"smart quotes" instead of "straight quotes"19Chatbot artifacts"I hope this helps!", "Let me know if..."20Cutoff disclaimers"As of my last training...", "While details are limited..."21Sycophantic tone"Great question!", "You're absolutely right!"22Filler phrases"In order to", "Due to the fact that"23Excessive hedging"could potentially possibly", "might arguably"24Generic conclusions"The future looks bright", "Exciting times lie ahead"

### Tier 1 (Dead Giveaways - NEVER USE)

delve, tapestry, vibrant, crucial, comprehensive, meticulous, embark, robust, seamless, groundbreaking, leverage, synergy, transformative, paramount, multifaceted, myriad, cornerstone, reimagine, empower, catalyst, invaluable, bustling, nestled, realm

### Tier 2 (Suspicious - Use Sparingly)

furthermore, moreover, paradigm, holistic, utilize, facilitate, nuanced, illuminate, encompasses, catalyze, proactive, ubiquitous, quintessential

### Tier 3 (Context-Dependent - Watch Density)

landscape (abstract), journey (metaphorical), ecosystem, framework, roadmap, touchpoint, pain point, streamline, optimize, scalable

See references/vocabulary.md for the complete 500+ term database.

### Statistical Signals

Check these when doing comprehensive analysis:

SignalHumanAIWhyBurstiness0.5-1.00.1-0.3Humans write in bursts; AI is metronomicType-token ratio0.5-0.70.3-0.5AI reuses vocabularySentence length CoVHighLowAI makes same-length sentencesTrigram repetition<0.05>0.10AI reuses 3-word phrases

See references/statistical-signals.md for formulas and implementation.

### Personality Injection (NEW)

When asked to add human touches or make text sound less perfect:

### 1. Parenthetical Asides

Add authentic reactions in parentheses:

(honestly, this part gets me every time)
(which, let's be real, is kind of absurd)
(I still don't fully understand why)

Rule: Use for genuine reactions, not filler. 1-3 per 500 words max.

### 2. Strategic Misspellings

Natural typos that don't hurt credibility:

Acceptable: "teh" → "the", "recieve" → "receive", "definately" → "definitely"
Avoid: Professional contexts, names, technical terms, key claims

Rule: 1-2 per 1000 words in casual contexts only.

### 3. Tangents and Going Off Script

Let thoughts wander naturally:

"Speaking of which..."
"That reminds me..."
"Okay, slight tangent..."
"Wait, this connects to..."

Rule: Allow 1-2 per longer piece (1000+ words).

### 4. Random Thoughts

Inject authentic human processing:

"Honestly didn't think this would work but..."
"I keep coming back to..."
"There's something unsettling about..."
"I genuinely don't know how to feel about this"

Rule: Use when genuine uncertainty or strong reaction fits.

See references/personality-injection.md for comprehensive examples and context guidance.

### Process

Read the input text carefully
Identify all pattern instances (use the 24-pattern checklist)
Flag vocabulary issues (Tier 1 = remove, Tier 2 = reduce, Tier 3 = watch density)
Check statistics (if comprehensive analysis requested)
Rewrite problematic sections
Add personality (if requested or if text is sterile)
Verify naturally (read aloud, check rhythm)
Present humanized version with brief change summary

### Before (AI-sounding)

Great question! Here is an overview of sustainable energy. Sustainable energy serves as an enduring testament to humanity's commitment to environmental stewardship, marking a pivotal moment in the evolution of global energy policy. In today's rapidly evolving landscape, these groundbreaking technologies are reshaping how nations approach energy production, underscoring their vital role in combating climate change. The future looks bright. I hope this helps!

### After (Humanized)

Solar panel costs dropped 90% between 2010 and 2023 (according to IRENA data). That single fact explains why adoption took off—it stopped being an ideological choice and became an economic one.

Germany gets 46% of its electricity from renewables now. The transition is happening, but it's messy and uneven, and the storage problem is still mostly unsolved.

### Changes Made

Removed "Great question!" and "I hope this helps!" (chatbot artifacts)
Removed "serves as an enduring testament" (significance inflation)
Removed "marking a pivotal moment" (AI vocabulary)
Removed "rapidly evolving landscape" (AI vocabulary)
Removed "groundbreaking", "underscoring", "vital" (AI vocabulary)
Removed "The future looks bright" (generic conclusion)
Added specific data points and sources
Added personal observation (parenthetical)
Added acknowledgment of complexity
Varied sentence rhythm

### Using the CLI Tool

The scripts/humanize.js tool provides command-line access:

# Score text (0-100, higher = more AI-like)
node scripts/humanize.js score "Your text here"

# Full analysis report
node scripts/humanize.js analyze -f draft.md

# Humanization suggestions
node scripts/humanize.js suggest article.txt

# Auto-fix common patterns
node scripts/humanize.js fix --autofix -f article.txt

### Always-On Mode

To make this skill's rules your default writing style (not just when explicitly asked to humanize):

Ban Tier 1 vocabulary completely from all writing
Kill filler phrases automatically
No sycophancy, chatbot artifacts, or generic conclusions
Vary sentence length, have opinions, use concrete specifics
If you wouldn't say it in conversation, don't write it

These rules can be added to SOUL.md or agent personality config for permanent application.

### Tips for Effective Humanization

Start with Tier 1 vocabulary → Easiest wins, most obvious AI tells
Check for em dashes → Quick scan for — overuse
Read the last paragraph → Generic conclusions are common
Listen to rhythm → Read aloud, hear if it sounds robotic
Don't over-polish → Some imperfection is human
Match the context → Formal documents need different treatment than social posts
Preserve meaning → Remove patterns without changing core message

### Troubleshooting

Text still sounds robotic after fixing patterns?
→ You removed AI tells but didn't add personality. Review references/personality-injection.md.

Humanization made text too casual?
→ Personality injection should match context. Use fewer asides/typos in formal writing.

Text is TOO perfect now?
→ Add strategic imperfection: vary sentence length, include a tangent, acknowledge uncertainty.

Not sure if a word is AI vocabulary?
→ Check references/vocabulary.md. When in doubt, ask "Would I say this in conversation?"
## Trust
- Source: tencent
- Verification: Indexed source record
- Publisher: Kevjade
- Version: 1.0.0
## Source health
- Status: healthy
- Item download looks usable.
- Yavira can redirect you to the upstream package for this item.
- Health scope: item
- Reason: direct_download_ok
- Checked at: 2026-05-03T00:45:11.942Z
- Expires at: 2026-05-10T00:45:11.942Z
- Recommended action: Download for OpenClaw
## Links
- [Detail page](https://openagent3.xyz/skills/operator-humanizer)
- [Send to Agent page](https://openagent3.xyz/skills/operator-humanizer/agent)
- [JSON manifest](https://openagent3.xyz/skills/operator-humanizer/agent.json)
- [Markdown brief](https://openagent3.xyz/skills/operator-humanizer/agent.md)
- [Download page](https://openagent3.xyz/downloads/operator-humanizer)