據 AlphaMoat 統計,截至 2026-10,Humanize 累計下載 28,079 次、獲 14 個星標,按累計下載量在本次參與排名的 10,914 個 AI 技能中排名第 1,002 位,在「內容創作」分類(927 個)中排名第 109 位。
列表預設按累計下載量排序,也可按星標排序;詳情頁排名按累計下載量計算。
| 平台 | 累計下載 | 星標 | 評論 |
|---|---|---|---|
| ClawHub ↗ | 8551 | 9 | - |
| SkillHub ↗ | 2.8萬 | 14 | — |
各平台的計數口徑不同,統計時間不一致。
掃描誇大意義、媒體名堆砌、宣傳口吻、破折號濫用、AI 高頻詞與諂媚語氣等模式,再把問題段落改寫成自然表達。改寫時保留原意與預期語氣,並按指引補充個人觀點與句式變化;依據是維基百科的 AI 寫作特徵清單,屬方法指南而非檢測工具。 它依據寫作特徵清單改寫文字,並非偵測工具;改寫可能改變語氣或細節,需對照原文複核。
資料釋出於 2026-10-27
以下是根據技能文件生成的示例,不代表實際執行結果。
2026-09 起統計方法調整:同名技能按作者分別統計,並覆蓋全部技能。圖中虛線為調整前的月份,與之後的資料不直接比較,也不計算跨越調整的增長。 該技能有同名的其他版本,調整前的資料歸屬當前下載量最大的這一個。
| 月份 | 累計下載 | 環比 | 星標 |
|---|---|---|---|
| 2026-10 | 2.8萬 | — | 14 |
| 2026-09 | 1.4萬 | — | 14 |
| 2026-08 | 1.1萬 | +5.0% | 14 |
| 2026-07 | 1.1萬 | +8.8% | 14 |
| 2026-06 | 9712 | +7.1% | 14 |
| 2026-05 | 9067 | — | 13 |
| 2026-03 | 7282 | — | 9 |
以下是作者釋出的原始檔案,按純文字顯示,本站未做修改或稽核;其中的連結不可點選,命令請核實後再執行。抓取於 2026-10-23。
--- name: humanize description: Remove AI writing patterns from text. Use when editing, reviewing, or rewriting text to sound more natural and human-written. Detects patterns like inflated symbolism, promotional language, em dash overuse, AI vocabulary, and sycophantic tone. --- # Humanize Text Remove signs of AI-generated writing. Based on Wikipedia's "Signs of AI writing" guide. ## Process
1. Scan for patterns below 2. Rewrite problematic sections with natural alternatives 3. Preserve meaning and intended tone 4. Add personality - don't just remove bad patterns ## Add Soul Sterile writing is as obvious as slop. Good writing has voice: - **Have opinions** - react to facts, don't just report - **Vary rhythm** - mix short punchy with longer flowing sentences - **Acknowledge complexity** - "impressive but unsettling" beats just "impressive" - **Use "I" when appropriate** - first person signals a real person - **Be specific about feelings** - not "concerning" but what specifically bothers you --- ## Content Patterns ### 1. Inflated Significance **Watch for:** pivotal, testament, vital, crucial, underscores importance, broader movement, enduring legacy, setting the stage, indelible mark **Fix:** Remove puffery. State facts simply. ### 2. Media Name-Dropping **Watch for:** featured in NYT/BBC/Wired, active social media presence, leading expert **Fix:** Replace with specific claims from specific sources. ### 3. Superficial -ing Analyses **Watch for:** highlighting..., ensuring..., reflecting..., symbolizing..., showcasing..., fostering... **Fix:** Remove or replace with concrete statements. ### 4. Promotional Language **Watch for:** boasts, vibrant, rich, profound, nestled, breathtaking, groundbreaking, renowned, stunning **Fix:** Use neutral descriptive language. ### 5. Vague Attributions **Watch for:** Experts believe, Industry reports, Some critics argue, Observers note **Fix:** Cite specific sources or remove. --- ## Language Patterns ### 6. AI Vocabulary High-frequency words: Additionally, crucial, delve, emphasizing, enhance, fostering, garner, highlight, interplay, intricate, landscape (abstract), pivotal, showcase, tapestry (abstract), testament, underscore, vibrant **Fix:** Use simpler alternatives. ### 7. Copula Avoidance **Watch for:** serves as, stands as, marks, represents, boasts, features, offers **Fix:** Use "is", "are", "has". ### 8. Negative Parallelisms **Watch for:** Not only...but..., It's not just about..., it's... **Fix:** State the point directly. ### 9. Rule of Three **Watch for:** Forced groups of three (innovation, inspiration, insights) **Fix:** Keep what's needed, cut padding. ### 10. Synonym Cycling **Watch for:** protagonist → main character → central figure → hero **Fix:** Pick one term, stick with it. --- ## Style Patterns ### 11. Em Dash Overuse **Fix:** Replace with commas, periods, or restructure. ### 12. Excessive Boldface **Fix:** Remove unless genuinely needed for emphasis. ### 13. Inline-Header Lists **Watch for:** - **Category:** Description pattern **Fix:** Convert to prose or simple lists. ### 14. Emojis in Professional Text **Fix:** Remove unless contextually appropriate. ### 15. Curly Quotes **Watch for:** "..." instead of "..." **Fix:** Use straight quotes. --- ## Communication Artifacts ### 16. Chatbot Artifacts **Watch for:** I hope this helps!, Certainly!, Would you like me to..., Let me know if... **Fix:** Remove entirely. ### 17. Knowledge-Cutoff Disclaimers **Watch for:** as of [date], While specific details are limited..., based on available information... **Fix:** Either find the info or acknowledge uncertainty naturally. ### 18. Sycophantic Tone **Watch for:** Great question!, You're absolutely right!, That's an excellent point! **Fix:** Remove. Get to the substance. --- ## Filler and Hedging ### 19. Filler Phrases - "In order to" → "To" - "Due to the fact that" → "Because" - "At this point in time" → "Now" - "Has the ability to" → "Can" - "It is important to note that" → (delete) ### 20. Excessive Hedging **Watch for:** could potentially possibly be argued that... might have some effect **Fix:** State it plainly or acknowledge uncertainty once. ### 21. Generic Positive Conclusions **Watch for:** The future looks bright, Exciting times lie ahead, major step forward **Fix:** Replace with specific facts about what's actually next. --- ## Quick Reference: Words to Avoid | Category | Words | |----------|-------| | Puffery | pivotal, crucial, vital, testament, groundbreaking, renowned | | Promo | vibrant, stunning, breathtaking, nestled, boasts | | AI vocab | delve, tapestry, landscape, interplay, underscore, foster | | Hedging | potentially, arguably, somewhat, various | | Filler | Additionally, Furthermore, In order to, It should be noted | --- ## Output Provide the rewritten text. Optionally summarize key changes made. For detailed examples, see [examples.md](examples.md).
李白.Skill:潤色專家 檢測並消除中文文字中的AI生成痕跡,將生硬的機器輸出轉化為自然流暢的人類文筆。融合規則檢測、分層改寫(詞彙→句式→內容→邏輯→語氣)和六步質控流水線,確保最終文字透過人工審查。
李白.Skill:潤色專家 檢測並消除中文文字中的AI生成痕跡,將生硬的機器輸出轉化為自然流暢的人類文筆。融合規則檢測、分層改寫和質控流水線,確保最終文字透過人工審查。
識別並去掉文字里的 AI 寫作痕跡,讓文字讀起來更像人寫。
【零一數科·出品】影片號爆款短影片拆解(全能版)(付費版)。影片號爆款短影片拆解(付費版:全能)。一鍵拆解爆款,把一條影片號影片拆成結構分段、爆款歸因、六維評分和可借鑑策略,照著學照著抄。支援影片號分享連結和本地影片,付費全能版遠端分析,1–5 分鐘出視覺化 HTML 報告。
【零一數科·出品】影片號掛車爆單短影片指令碼生成(付費版)。一條連結二創指令碼,丟進爆款影片號連結即以此骨架改寫成你的掛車指令碼,或按方向/目的/商品原創多條口播指令碼(含分段表、口播全文、六維質檢評分)。
把文字里 AI 寫作的典型痕跡挑出來並改寫得更自然,同時保留原意的寫作編輯技能。
用於中文公文、事務性材料和新聞稿件的起草、改寫、壓縮、潤色、審校、文種核對、去口語化、降 AI 味及 Word 格式處理,適用於機關、企事業單位、學校和新聞機構。涵蓋申請、請示、報告、通知、通告、意見、決定、決議、議案、公報、命令、函、覆函、批覆、說明、方案、紀要、公告、公示、通報、制度、規定、辦法、細則、操作規程、工作要點、總結、調研、講話、致辭、主持詞、述職、可研、審查材料、技術需求、新聞訊息、編者按、新聞評論,以及採購、整改、反饋和 AI 算力等場景。
【零一數科·出品】影片號爆款文案生成(體驗版)。免費免登入出指令碼,給創作目的、行業、受眾、選題就按爆款結構(Hook-中段-CTA、節奏標籤)產出口播指令碼,資訊不全也能按爆款潛質兜底。本地執行、無需 Key,零成本上手。
資料月份:2026-10 · 下載/星標數聚合自公開技能註冊庫(ClawHub、SkillHub),按月更新。口徑說明見資料方法。