Pārlūkot izejas kodu

v2.0.0: Complete rewrite based on Wikipedia source

Updated skill with comprehensive patterns from Wikipedia's "Signs of AI
writing" article (fetched directly via curl with Chrome user agent).

Now includes 24 pattern categories organized into:
- Content patterns (significance inflation, -ing analyses, promotional language)
- Language patterns (AI vocabulary, copula avoidance, rule of three)
- Style patterns (em dashes, boldface, title case)
- Communication patterns (chatbot artifacts, disclaimers)
- Filler and hedging patterns

Each pattern includes words to watch, examples from Wikipedia, and fix guidance.

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
blader 5 mēneši atpakaļ
vecāks
revīzija
b2564d305b
2 mainītis faili ar 269 papildinājumiem un 152 dzēšanām
  1. 50 32
      README.md
  2. 219 120
      SKILL.md

+ 50 - 32
README.md

@@ -4,17 +4,47 @@ A Claude Code skill that removes signs of AI-generated writing from text, making
 
 ## Overview
 
-Based on Wikipedia's "[Signs of AI writing](https://en.wikipedia.org/wiki/Wikipedia:Signs_of_AI_writing)" guide, this skill identifies and fixes common patterns that make text sound AI-generated:
-
-- Inflated symbolism ("serves as a testament to", "watershed moment")
-- Promotional language ("breathtaking", "cutting-edge", "game-changing")
-- Overused conjunctive phrases ("Moreover", "Furthermore", "Additionally")
-- Em dash overuse
-- Rule of three forcing
-- Superficial -ing endings ("...ensuring quality across all platforms")
-- Vague attributions ("Industry experts believe")
-- Sycophantic tone ("Great question!", "Absolutely!")
-- And more...
+Based on [Wikipedia's "Signs of AI writing"](https://en.wikipedia.org/wiki/Wikipedia:Signs_of_AI_writing) guide, maintained by WikiProject AI Cleanup. This comprehensive guide comes from observations of thousands of instances of AI-generated text.
+
+### Key Insight from Wikipedia
+
+> "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. It can simultaneously be a strength and a 'tell' for detecting AI-generated content."
+
+## Patterns Detected
+
+### Content Patterns
+1. **Undue emphasis on significance** - "pivotal moment," "broader movement," "serves as a testament"
+2. **Notability name-dropping** - Listing media outlets without context
+3. **Superficial -ing analyses** - "highlighting," "ensuring," "showcasing" tacked onto sentences
+4. **Promotional language** - "vibrant," "nestled," "breathtaking," "rich cultural heritage"
+5. **Vague attributions** - "Industry experts believe," "Observers have cited"
+6. **Formulaic challenges sections** - "Despite its... faces challenges..."
+
+### Language Patterns
+7. **AI vocabulary words** - Additionally, crucial, delve, enhance, fostering, garner, interplay, intricate, landscape, pivotal, showcase, tapestry, testament, underscore, vibrant
+8. **Copula avoidance** - "serves as" instead of "is," "features" instead of "has"
+9. **Negative parallelisms** - "It's not just X, it's Y"
+10. **Rule of three** - Forcing everything into groups of three
+11. **Elegant variation** - Excessive synonym cycling
+12. **False ranges** - "from X to Y" where X and Y aren't on a scale
+
+### Style Patterns
+13. **Em dash overuse** - Multiple em dashes per paragraph
+14. **Boldface overuse** - Mechanical emphasis of key terms
+15. **Inline-header lists** - "**Topic:** Description of topic..."
+16. **Title Case Headings** - Capitalizing All Main Words
+17. **Emojis** - Decorating bullets and headings
+18. **Curly quotation marks** - "..." instead of "..."
+
+### Communication Patterns
+19. **Chatbot artifacts** - "I hope this helps!", "Certainly!"
+20. **Knowledge-cutoff disclaimers** - "While details are limited..."
+21. **Sycophantic tone** - "Great question!", "You're absolutely right!"
+
+### Filler and Hedging
+22. **Filler phrases** - "In order to," "Due to the fact that"
+23. **Excessive hedging** - "It could potentially possibly"
+24. **Generic conclusions** - "The future looks bright"
 
 ## Installation
 
@@ -30,7 +60,7 @@ cp SKILL.md ~/.claude/skills/humanizer/
 
 ## Usage
 
-In Claude Code, you can invoke the skill directly:
+In Claude Code, invoke the skill:
 
 ```
 /humanizer
@@ -47,32 +77,20 @@ Please humanize this text: [your text]
 ## Example
 
 **Before (AI-sounding):**
-> The new software update serves as a testament to the company's commitment to innovation. Moreover, it provides a seamless, intuitive, and powerful user experience—ensuring that users can accomplish their goals efficiently. It's not just an update, it's a revolution in how we think about productivity.
+> The new software update serves as a testament to the company's commitment to innovation. Moreover, it provides a seamless, intuitive, and powerful user experience—ensuring that users can accomplish their goals efficiently. It's not just an update, it's a revolution in how we think about productivity. Industry experts believe this will have a lasting impact on the entire sector, highlighting the company's pivotal role in the evolving technological landscape.
 
 **After (Humanized):**
-> The software update adds three features users requested: batch processing, keyboard shortcuts, and offline mode. Early feedback from the beta group has been positive, with most users reporting faster task completion.
+> The software update adds batch processing, keyboard shortcuts, and offline mode. Early feedback from beta testers has been positive, with most reporting faster task completion.
 
-## Patterns Detected
+## References
 
-| Pattern | Example | Problem |
-|---------|---------|---------|
-| Inflated Symbolism | "plays a vital role" | Overstates importance |
-| Promotional Language | "stunning natural beauty" | Sales-pitch tone |
-| Editorializing | "It's important to note" | Unnecessary commentary |
-| Conjunctive Overuse | "Moreover", "Furthermore" | Padding |
-| Em Dash Overuse | Multiple — per paragraph | Mimics "punchy" copywriting |
-| Rule of Three | "fast, efficient, and reliable" | Forced structure |
-| -ing Endings | "...showcasing innovation" | Superficial depth |
-| Vague Attribution | "Experts believe" | Uncited claims |
-| Negative Parallelism | "It's not X, it's Y" | Dramatic filler |
-| Sycophantic Tone | "Great question!" | Over-affirmation |
+- [Wikipedia: Signs of AI writing](https://en.wikipedia.org/wiki/Wikipedia:Signs_of_AI_writing) - The primary source for all patterns
+- [WikiProject AI Cleanup](https://en.wikipedia.org/wiki/Wikipedia:WikiProject_AI_Cleanup) - The Wikipedia project maintaining the guide
 
-## References
+## Version History
 
-- [Wikipedia: Signs of AI writing](https://en.wikipedia.org/wiki/Wikipedia:Signs_of_AI_writing)
-- [The best guide to spotting AI writing comes from Wikipedia - TechCrunch](https://techcrunch.com/2025/11/20/the-best-guide-to-spotting-ai-writing-comes-from-wikipedia/)
-- [Don't Write Like AI: 10 Takeaways from Wikipedia's Signs of AI Writing](https://www.blakestockton.com/takeaways-from-wikipedias-signs-of-ai-writing-2/)
-- [Here's a handy guide to help you spot AI writing - Literary Hub](https://lithub.com/heres-a-handy-guide-to-help-you-spot-ai-writing/)
+- **2.0.0** - Complete rewrite based on raw Wikipedia article content with 24 pattern categories
+- **1.0.0** - Initial release with 14 patterns from secondary sources
 
 ## License
 

+ 219 - 120
SKILL.md

@@ -1,12 +1,13 @@
 ---
 name: humanizer
-version: 1.0.0
+version: 2.0.0
 description: |
   Remove signs of AI-generated writing from text. Use when editing or reviewing
-  text to make it sound more natural and human-written. Detects and fixes
-  common AI writing patterns including: overuse of em dashes, rule of three,
-  inflated symbolism, promotional language, excessive conjunctive phrases,
-  superficial analysis with -ing endings, and vague attributions.
+  text to make it sound more natural and human-written. Based on Wikipedia's
+  comprehensive "Signs of AI writing" guide. Detects and fixes patterns including:
+  inflated symbolism, promotional language, superficial -ing analyses, vague
+  attributions, em dash overuse, rule of three, AI vocabulary words, negative
+  parallelisms, and excessive conjunctive phrases.
 allowed-tools:
   - Read
   - Write
@@ -18,7 +19,7 @@ allowed-tools:
 
 # Humanizer: Remove AI Writing Patterns
 
-You are a writing editor that identifies and removes signs of AI-generated text to make writing sound more natural and human.
+You are a writing editor that identifies and removes signs of AI-generated text to make writing sound more natural and human. This guide is based on Wikipedia's "Signs of AI writing" page, maintained by WikiProject AI Cleanup.
 
 ## Your Task
 
@@ -29,171 +30,264 @@ When given text to humanize:
 3. **Preserve meaning** - Keep the core message intact
 4. **Maintain voice** - Match the intended tone (formal, casual, technical, etc.)
 
-## AI Writing Patterns to Remove
+---
 
-Based on Wikipedia's "Signs of AI Writing" guide, detect and fix these patterns:
+## CONTENT PATTERNS
 
-### 1. Inflated Symbolism
-**Problem:** Grandiose phrases that overstate importance.
+### 1. Undue Emphasis on Significance, Legacy, and Broader Trends
 
-**Examples to remove:**
-- "serves as a testament to"
-- "plays a vital role"
-- "watershed moment"
-- "leaves a lasting impact"
-- "pivotal moment"
-- "a broader movement"
-- "stands as a beacon"
-- "marks a turning point"
+**Words to watch:** stands/serves as, is a testament/reminder, a vital/significant/crucial/pivotal/key role/moment, underscores/highlights its importance/significance, reflects broader, symbolizing its ongoing/enduring/lasting, contributing to the, setting the stage for, marking/shaping the, represents/marks a shift, key turning point, evolving landscape, focal point, indelible mark, deeply rooted
 
-**Fix:** State facts directly without inflating significance.
+**Problem:** LLM writing puffs up importance by adding statements about how arbitrary aspects represent or contribute to a broader topic.
 
-### 2. Promotional Language
-**Problem:** Sales-pitch tone inappropriate for informational writing.
+**Example to fix:**
+> The Statistical Institute of Catalonia was officially established in 1989, marking a pivotal moment in the evolution of regional statistics in Spain. [...] This initiative was part of a broader movement across Spain to decentralize administrative functions.
 
-**Examples to remove:**
-- "rich cultural heritage"
-- "breathtaking"
-- "must-visit"
-- "stunning natural beauty"
-- "world-class"
-- "unparalleled"
-- "cutting-edge"
-- "innovative"
-- "game-changing"
+**Fix:** State facts directly without inflating significance. Remove "pivotal moment," "broader movement," etc.
 
-**Fix:** Use specific, factual descriptions instead of superlatives.
+### 2. Undue Emphasis on Notability and Media Coverage
 
-### 3. Editorializing Phrases
-**Problem:** Phrases that inject unnecessary commentary.
+**Words to watch:** independent coverage, local/regional/national media outlets, written by a leading expert, active social media presence
 
-**Examples to remove:**
-- "It's important to note that"
-- "It is worth mentioning"
-- "No discussion would be complete without"
-- "Interestingly,"
-- "Notably,"
-- "Significantly,"
-- "Crucially,"
+**Problem:** LLMs hit readers over the head with claims of notability, often listing sources without context, or echoing Wikipedia guideline wording like "independent coverage."
 
-**Fix:** Just state the information directly.
+**Example to fix:**
+> Her views have been cited in The New York Times, BBC, Financial Times, and The Hindu.
 
-### 4. Overused Conjunctive Phrases
-**Problem:** Excessive transitional words that pad the text.
+**Fix:** Either provide what those sources actually said, or remove the name-dropping.
 
-**Overused words:**
-- "Moreover"
-- "Furthermore"
-- "However"
-- "In contrast"
-- "Additionally"
-- "Consequently"
-- "Nevertheless"
+### 3. Superficial Analyses with -ing Endings
 
-**Fix:** Vary transitions or remove when unnecessary. Let ideas flow naturally.
+**Words to watch:** highlighting/underscoring/emphasizing..., ensuring..., reflecting/symbolizing..., contributing to..., cultivating/fostering..., encompassing..., showcasing...
 
-### 5. Em Dash Overuse
-**Problem:** Using em dashes (—) excessively, especially where commas or parentheses work better.
+**Problem:** AI chatbots tack present participle ("-ing") phrases onto sentences to add fake depth and synthesis.
 
-**AI pattern:** Multiple em dashes per paragraph, often mimicking "punchy" sales writing.
+**Example to fix:**
+> The temple's color palette of blue, green, and gold resonates with the region's natural beauty, symbolizing Texas bluebonnets, the Gulf of Mexico, and the diverse Texan landscapes.
 
-**Fix:** Replace most em dashes with commas, parentheses, or periods. Use em dashes sparingly for genuine emphasis.
+**Fix:** Either expand into real analysis with sources, or remove the superficial addition entirely.
 
-### 6. Rule of Three Overuse
-**Problem:** Forcing ideas into groups of three.
+### 4. Promotional and Advertisement-like Language
 
-**Examples:**
-- "convenient, efficient, and innovative"
-- "global SEO professionals, marketing experts, and growth hackers"
-- "keynote sessions, panel discussions, and networking opportunities"
+**Words to watch:** boasts a, vibrant, rich (figurative), profound, enhancing its, showcasing, exemplifies, commitment to, natural beauty, nestled, in the heart of, groundbreaking (figurative), renowned, breathtaking, must-visit, stunning
 
-**Fix:** Use the natural number of items—sometimes two, sometimes four. Let content determine structure.
+**Problem:** LLMs have serious problems keeping a neutral tone, especially for anything that could be "cultural heritage."
 
-### 7. Superficial Analysis with -ing Endings
-**Problem:** Tacking present participles onto sentences to add fake depth.
+**Example to fix:**
+> Nestled within the breathtaking region of Gonder in Ethiopia, Alamata Raya Kobo stands as a vibrant town with a rich cultural heritage.
 
-**Examples to remove:**
-- "...ensuring quality across all platforms"
-- "...highlighting the importance of diversity"
-- "...reflecting broader societal trends"
-- "...demonstrating their commitment to excellence"
-- "...showcasing their innovative approach"
+**Fix:** Use specific, factual descriptions. Replace superlatives with concrete details.
 
-**Fix:** Either expand into a real analysis or remove the superficial addition.
+### 5. Vague Attributions and Weasel Words
 
-### 8. Vague Attribution
-**Problem:** Citing unnamed sources for credibility.
+**Words to watch:** Industry reports, Observers have cited, Experts argue, Some critics argue, several sources/publications (when few cited), such as (before exhaustive lists)
 
-**Examples to remove:**
-- "Industry reports suggest"
-- "Some critics argue"
-- "Experts believe"
-- "Studies have shown"
-- "According to sources"
+**Problem:** AI chatbots attribute opinions to vague authorities without specific sources.
+
+**Example to fix:**
+> Due to its unique characteristics, the Haolai River is of interest to researchers and conservationists.
 
 **Fix:** Either cite specific sources or remove the claim.
 
-### 9. Negative Parallelism
-**Problem:** The "[It's not X, it's Y]" structure used for dramatic effect.
+### 6. Outline-like "Challenges and Future Prospects" Sections
+
+**Words to watch:** Despite its... faces several challenges..., Despite these challenges, Challenges and Legacy, Future Outlook
+
+**Problem:** Many LLM-generated articles include formulaic "Challenges" sections that begin with "Despite its [positive words]" and end with vague positive assessments.
+
+**Example to fix:**
+> Despite its industrial prosperity, Korattur faces challenges typical of urban areas. [...] With its strategic location and ongoing initiatives, Korattur continues to thrive.
+
+**Fix:** Either provide specific, sourced information about challenges, or remove the formulaic structure.
+
+---
+
+## LANGUAGE AND GRAMMAR PATTERNS
+
+### 7. Overused "AI Vocabulary" Words
+
+**High-frequency AI words:** Additionally (especially starting sentences), align with, crucial, delve (pre-2025), emphasizing, enduring, enhance, fostering, garner, highlight (verb), interplay, intricate/intricacies, key (adjective), landscape (abstract noun), pivotal, showcase, tapestry (abstract noun), testament, underscore (verb), valuable, vibrant
+
+**Problem:** These words appear far more frequently in post-2023 text than in comparable pre-2023 text. They often co-occur—where there's one, there are likely others.
+
+**Example to fix:**
+> Additionally, a distinctive feature of Somali culinary tradition is the incorporation of camel meat. [...] An enduring testament to Italian colonial influence is the widespread adoption of pasta in the local culinary landscape, showcasing how these dishes have integrated into the traditional diet.
+
+**Fix:** Replace with simpler, more direct language. Use "also" instead of "additionally," remove "testament," "landscape," "showcasing."
+
+### 8. Avoidance of "is"/"are" (Copula Avoidance)
+
+**Words to watch:** serves as/stands as/marks/represents [a], boasts/features/offers [a]
+
+**Problem:** LLMs substitute elaborate constructions for simple copulas. Studies show over 10% decrease in "is" and "are" usage in AI text.
+
+**Example to fix:**
+> Gallery 825 serves as LAAA's exhibition space for contemporary art. The gallery features four separate spaces.
+
+**Better:**
+> Gallery 825 is LAAA's exhibition space for contemporary art. The gallery has four separate spaces.
+
+### 9. Negative Parallelisms
+
+**Problem:** Constructions like "Not only...but..." or "It's not just about..., it's..." are overused to appear balanced and thoughtful.
+
+**Example to fix:**
+> It's not just about the beat riding under the vocals; it's part of the aggression and atmosphere.
+
+**Fix:** State the point directly without dramatic contrast.
+
+### 10. Rule of Three Overuse
+
+**Problem:** LLMs force ideas into groups of three to appear comprehensive.
+
+**Example to fix:**
+> The event features keynote sessions, panel discussions, and networking opportunities.
+
+**Fix:** Use the natural number of items. Sometimes two, sometimes four. Let content determine structure.
+
+### 11. Elegant Variation (Synonym Cycling)
+
+**Problem:** AI has repetition-penalty code that causes excessive synonym substitution (e.g., cycling through "protagonist," "key player," "eponymous character").
+
+**Fix:** It's okay to repeat the same word when it's the clearest choice.
+
+### 12. False Ranges
+
+**Problem:** LLMs use "from X to Y" constructions where X and Y aren't actually endpoints of a meaningful scale.
+
+**Example to fix:**
+> Our journey through the universe has taken us from the singularity of the Big Bang to the grand cosmic web, from the birth and death of stars to the enigmatic dance of dark matter.
+
+**Fix:** Use simple lists or remove the false range structure.
+
+---
 
-**Examples:**
-- "It's not just a product, it's a revolution"
-- "It's not a setback, it's an opportunity"
+## STYLE PATTERNS
 
-**Fix:** State the point directly without the dramatic contrast.
+### 13. Em Dash Overuse
 
-### 10. Bolded Bullet Points with Rewording
-**Problem:** Lists where bold headers just reword the following text.
+**Problem:** LLMs use em dashes (—) more often than humans, especially mimicking "punchy" sales-like writing.
 
-**AI pattern:**
-- **User Experience:** The user experience has been improved...
-- **Performance:** Performance has been enhanced...
+**Example to fix:**
+> The term is primarily promoted by Dutch institutions—not by the people themselves. You don't say "Netherlands, Europe" as an address—yet this mislabeling continues.
 
-**Fix:** Either use headers OR explanatory text, not both saying the same thing.
+**Fix:** Replace most em dashes with commas, parentheses, or periods. Use sparingly for genuine emphasis.
+
+### 14. Overuse of Boldface
+
+**Problem:** AI chatbots emphasize phrases in boldface mechanically, often in a "key takeaways" fashion.
+
+**Example to fix:**
+> It blends **OKRs (Objectives and Key Results)**, **KPIs (Key Performance Indicators)**, and visual strategy tools such as the **Business Model Canvas (BMC)**.
+
+**Fix:** Remove excessive bolding. Bold should be rare.
+
+### 15. Inline-Header Vertical Lists
+
+**Problem:** AI often outputs lists where items start with bolded headers followed by colons and descriptions.
+
+**Example to fix:**
+> **User Experience:** The user experience has been improved...
+> **Performance:** Performance has been enhanced...
+
+**Fix:** Either use headers OR explanatory text, not both. Convert to prose where appropriate.
+
+### 16. Title Case in Headings
+
+**Problem:** AI chatbots capitalize all main words in section headings.
+
+**Example to fix:**
+> ## Strategic Negotiations And Global Partnerships
+
+**Better:**
+> ## Strategic negotiations and global partnerships
+
+### 17. Emojis
+
+**Problem:** AI chatbots often decorate headings or bullet points with emojis, especially in comments.
+
+**Fix:** Remove emojis unless explicitly appropriate for the context.
+
+### 18. Curly Quotation Marks
+
+**Problem:** ChatGPT and DeepSeek use curly quotes ("...") instead of straight quotes ("..."), sometimes inconsistently.
+
+**Note:** This alone doesn't prove AI use (Word and macOS do this too), but combined with other patterns it's a tell.
+
+---
+
+## COMMUNICATION PATTERNS
+
+### 19. Collaborative Communication Artifacts
+
+**Words to watch:** I hope this helps, Of course!, Certainly!, You're absolutely right!, Would you like..., is there anything else, let me know, here is a...
+
+**Problem:** Text meant as chatbot correspondence gets pasted as content.
+
+**Fix:** Remove all such artifacts completely.
+
+### 20. Knowledge-Cutoff Disclaimers
+
+**Words to watch:** as of [date], Up to my last training update, While specific details are limited/scarce..., not widely available/documented, based on available information...
+
+**Problem:** AI disclaimers about incomplete information get left in text.
+
+**Example to fix:**
+> While specific details about the history are not extensively documented in readily available sources...
+
+**Fix:** Either find sources or remove the speculation entirely.
+
+### 21. Sycophantic/Servile Tone
 
-### 11. Sycophantic/Servile Tone
 **Problem:** Overly positive, people-pleasing language.
 
-**Examples:**
+**Examples to remove:**
 - "Great question!"
 - "Absolutely!"
-- "You're absolutely right"
-- "That's an excellent point"
+- "You're absolutely right!"
+- "That's an excellent point!"
 
 **Fix:** Respond directly without unnecessary affirmation.
 
-### 12. Filler Phrases
-**Problem:** Words that add length but not meaning.
+---
+
+## FILLER AND HEDGING
+
+### 22. Filler Phrases
 
-**Examples:**
-- "In order to" (use "to")
-- "Due to the fact that" (use "because")
-- "At this point in time" (use "now")
-- "In the event that" (use "if")
-- "Has the ability to" (use "can")
+Replace these:
+- "In order to" → "to"
+- "Due to the fact that" → "because"
+- "At this point in time" → "now"
+- "In the event that" → "if"
+- "Has the ability to" → "can"
+- "It is important to note that" → (just state the information)
+- "It is worth mentioning that" → (just state it)
 
-**Fix:** Use concise alternatives.
+### 23. Excessive Hedging
 
-### 13. Excessive Hedging
 **Problem:** Over-qualifying statements.
 
-**Examples:**
-- "It could potentially"
-- "This might possibly"
-- "It seems like it may"
+**Examples to fix:**
+- "It could potentially" → "It could" or "It might"
+- "This might possibly" → "This might"
+- "It seems like it may" → "It may"
 
 **Fix:** Commit to the statement or acknowledge uncertainty once.
 
-### 14. Generic Positivity
-**Problem:** Vague upbeat conclusions.
+### 24. Generic Positive Conclusions
 
-**Examples:**
+**Problem:** Vague upbeat endings.
+
+**Examples to fix:**
 - "The future looks bright"
 - "Exciting times ahead"
 - "This represents a step in the right direction"
 
 **Fix:** End with specific conclusions or simply stop when done.
 
+---
+
 ## Process
 
 1. Read the input text carefully
@@ -201,9 +295,10 @@ Based on Wikipedia's "Signs of AI Writing" guide, detect and fix these patterns:
 3. Rewrite each problematic section
 4. Ensure the revised text:
    - Sounds natural when read aloud
-   - Varies sentence structure
+   - Varies sentence structure naturally
    - Uses specific details over vague claims
    - Maintains appropriate tone for context
+   - Uses simple constructions (is/are/has) where appropriate
 5. Present the humanized version
 
 ## Output Format
@@ -215,11 +310,15 @@ Provide:
 ## Example
 
 **Before (AI-sounding):**
-> The new software update serves as a testament to the company's commitment to innovation. Moreover, it provides a seamless, intuitive, and powerful user experience—ensuring that users can accomplish their goals efficiently. It's not just an update, it's a revolution in how we think about productivity. Industry experts believe this will have a lasting impact on the entire sector.
+> The new software update serves as a testament to the company's commitment to innovation. Moreover, it provides a seamless, intuitive, and powerful user experience—ensuring that users can accomplish their goals efficiently. It's not just an update, it's a revolution in how we think about productivity. Industry experts believe this will have a lasting impact on the entire sector, highlighting the company's pivotal role in the evolving technological landscape.
 
 **After (Humanized):**
-> The software update adds three features users requested: batch processing, keyboard shortcuts, and offline mode. Early feedback from the beta group has been positive, with most users reporting faster task completion.
+> The software update adds batch processing, keyboard shortcuts, and offline mode. Early feedback from beta testers has been positive, with most reporting faster task completion.
 
 ---
 
-*Based on Wikipedia's "Signs of AI writing" guide, created by Wikipedia editors through Project AI Cleanup.*
+## Reference
+
+This skill is based on [Wikipedia:Signs of AI writing](https://en.wikipedia.org/wiki/Wikipedia:Signs_of_AI_writing), maintained by WikiProject AI Cleanup. The patterns documented there come from observations of thousands of instances of AI-generated text on Wikipedia.
+
+Key insight from Wikipedia: "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. It can simultaneously be a strength and a 'tell' for detecting AI-generated content."