{
  "schemaVersion": "1.0",
  "item": {
    "slug": "humanize-chinese",
    "name": "Humanize Chinese",
    "source": "tencent",
    "type": "skill",
    "category": "通讯协作",
    "sourceUrl": "https://clawhub.ai/swaylq/humanize-chinese",
    "canonicalUrl": "https://clawhub.ai/swaylq/humanize-chinese",
    "targetPlatform": "OpenClaw"
  },
  "install": {
    "downloadMode": "redirect",
    "downloadUrl": "/downloads/humanize-chinese",
    "sourceDownloadUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=humanize-chinese",
    "sourcePlatform": "tencent",
    "targetPlatform": "OpenClaw",
    "installMethod": "Manual import",
    "extraction": "Extract archive",
    "prerequisites": [
      "OpenClaw"
    ],
    "packageFormat": "ZIP package",
    "includedAssets": [
      "README.md",
      "SKILL.md",
      "evals/evals.json",
      "package.json",
      "scripts/compare_cn.py",
      "scripts/detect_cn.py"
    ],
    "primaryDoc": "SKILL.md",
    "quickSetup": [
      "Download the package from Yavira.",
      "Extract the archive and review SKILL.md first.",
      "Import or place the package into your OpenClaw setup."
    ],
    "agentAssist": {
      "summary": "Hand the extracted package to your coding agent with a concrete install brief instead of figuring it out manually.",
      "steps": [
        "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."
      ],
      "prompts": [
        {
          "label": "New install",
          "body": "I downloaded a skill package from Yavira. Read SKILL.md from the extracted folder and install it by following the included instructions. Then review README.md for any prerequisites, environment setup, or post-install checks. Tell me what you changed and call out any manual steps you could not complete."
        },
        {
          "label": "Upgrade existing",
          "body": "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. Then review README.md for any prerequisites, environment setup, or post-install checks. Summarize what changed and any follow-up checks I should run."
        }
      ]
    },
    "sourceHealth": {
      "source": "tencent",
      "status": "healthy",
      "reason": "direct_download_ok",
      "recommendedAction": "download",
      "checkedAt": "2026-04-23T16:43:11.935Z",
      "expiresAt": "2026-04-30T16:43:11.935Z",
      "httpStatus": 200,
      "finalUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=4claw-imageboard",
      "contentType": "application/zip",
      "probeMethod": "head",
      "details": {
        "probeUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=4claw-imageboard",
        "contentDisposition": "attachment; filename=\"4claw-imageboard-1.0.1.zip\"",
        "redirectLocation": null,
        "bodySnippet": null
      },
      "scope": "source",
      "summary": "Source download looks usable.",
      "detail": "Yavira can redirect you to the upstream package for this source.",
      "primaryActionLabel": "Download for OpenClaw",
      "primaryActionHref": "/downloads/humanize-chinese"
    },
    "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."
      ]
    },
    "downloadPageUrl": "https://openagent3.xyz/downloads/humanize-chinese",
    "agentPageUrl": "https://openagent3.xyz/skills/humanize-chinese/agent",
    "manifestUrl": "https://openagent3.xyz/skills/humanize-chinese/agent.json",
    "briefUrl": "https://openagent3.xyz/skills/humanize-chinese/agent.md"
  },
  "agentAssist": {
    "summary": "Hand the extracted package to your coding agent with a concrete install brief instead of figuring it out manually.",
    "steps": [
      "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."
    ],
    "prompts": [
      {
        "label": "New install",
        "body": "I downloaded a skill package from Yavira. Read SKILL.md from the extracted folder and install it by following the included instructions. Then review README.md for any prerequisites, environment setup, or post-install checks. Tell me what you changed and call out any manual steps you could not complete."
      },
      {
        "label": "Upgrade existing",
        "body": "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. Then review README.md for any prerequisites, environment setup, or post-install checks. Summarize what changed and any follow-up checks I should run."
      }
    ]
  },
  "documentation": {
    "source": "clawhub",
    "primaryDoc": "SKILL.md",
    "sections": [
      {
        "title": "Humanize Chinese AI Text v2.0",
        "body": "Comprehensive CLI for detecting and transforming Chinese AI-generated text. Makes robotic AI writing natural and human-like.\n\nv2.0 highlights: weighted 0-100 scoring, sentence-level analysis, sentence restructuring (merge/split), context-aware replacement, rhythm variation, vocabulary diversification, 7 style transforms, external pattern config (patterns_cn.json)."
      },
      {
        "title": "Quick Start",
        "body": "# Detect AI patterns (20+ categories, 0-100 score)\npython scripts/detect_cn.py text.txt\npython scripts/detect_cn.py text.txt -v          # verbose + worst sentences\npython scripts/detect_cn.py text.txt -s           # score only\npython scripts/detect_cn.py text.txt -j           # JSON output\n\n# Humanize text\npython scripts/humanize_cn.py text.txt -o clean.txt\npython scripts/humanize_cn.py text.txt --scene social\npython scripts/humanize_cn.py text.txt --scene tech -a   # aggressive mode\npython scripts/humanize_cn.py text.txt --seed 42         # reproducible\n\n# Apply writing styles\npython scripts/style_cn.py text.txt --style zhihu -o zhihu.txt\npython scripts/style_cn.py text.txt --style xiaohongshu\npython scripts/style_cn.py --list\n\n# Compare before/after\npython scripts/compare_cn.py text.txt --scene tech -a\npython scripts/compare_cn.py text.txt -o clean.txt"
      },
      {
        "title": "Scoring",
        "body": "Weighted 0-100 score with 4 severity levels:\n\nScoreLevelMeaning0-24LOWLikely human-written25-49MEDIUMSome AI signals50-74HIGHProbably AI-generated75-100VERY HIGHAlmost certainly AI"
      },
      {
        "title": "Detection Categories",
        "body": "🔴 Critical (weight: 8)\n\nCategoryExamplesThree-Part Structure首先...其次...最后, 一方面...另一方面, 其一...其二...其三Mechanical Connectors值得注意的是, 综上所述, 不难发现, 归根结底, 由此可见Empty Grand Words赋能, 闭环, 数字化转型, 协同增效, 全方位, 多维度\n\n🟠 High Signal (weight: 4)\n\nCategoryExamplesAI High-Frequency Words助力, 彰显, 底层逻辑, 抓手, 触达, 沉淀, 复盘Filler Phrases值得一提的是, 众所周知, 毫无疑问Balanced Arguments虽然...但是...同时, 既有...也有...更有Template Sentences随着...的不断发展, 在当今...时代, 作为...的重要组成部分\n\n🟡 Medium Signal (weight: 2)\n\nCategoryExamplesHedging Language在一定程度上, 某种程度上, 通常情况下 (>5 occurrences)List AddictionExcessive numbered/bulleted listsPunctuation OveruseDense em dashes, semicolonsExcessive Rhetoric对偶/排比句过多\n\n⚪ Style Signal (weight: 1.5)\n\nCategoryDescriptionUniform ParagraphsLow CV in paragraph lengthsLow BurstinessMonotonous sentence lengthsEmotional FlatnessLack of emotional/personal expressionsRepetitive StartersSame sentence starters >3 timesLow EntropyLow character-level entropy (predictable text)"
      },
      {
        "title": "Sentence-Level Analysis",
        "body": "With -v (verbose) mode, the detector identifies the most AI-like sentences:\n\n── 最可疑句子 ──\n  1. [16分] 随着人工智能技术的不断发展，在当今数字化转型时代...\n     原因: 数字化转型, 深度融合, 模板: 随着.*?的(不断)?发展"
      },
      {
        "title": "Transforms (applied in order)",
        "body": "Structure cleanup — Remove three-part structure (首先/其次/最后)\nPhrase replacement — Context-aware replacement of AI phrases (regex patterns first, then plain text, longest-first matching)\nSentence merge — Merge overly short consecutive sentences\nSentence split — Split long sentences at natural breakpoints (但是/不过/同时)\nPunctuation normalization — Reduce excessive semicolons, em dashes\nVocabulary diversification — Replace repeated words (进行/实现/提供 etc.) with synonyms\nParagraph rhythm — Vary uniform paragraph lengths (merge short, split long)\nCasual injection — Add human expressions (scene-dependent)\nParagraph shortening — For social/chat scenes"
      },
      {
        "title": "Scenes",
        "body": "SceneCasualnessBest Forgeneral0.3Default, balancedsocial0.7Social media, short poststech0.3Tech blogs, tutorialsformal0.1Formal articles, reportschat0.8Conversations, messaging"
      },
      {
        "title": "Aggressive Mode (-a)",
        "body": "Adds +0.3 casualness, more colloquial expressions, stronger sentence restructuring. Typical score reduction: 60-80 points on heavily AI-generated text."
      },
      {
        "title": "Reproducibility",
        "body": "Use --seed N for reproducible results (same input + seed = same output)."
      },
      {
        "title": "Writing Style Transforms",
        "body": "7 specialized Chinese writing styles:\n\nStyleNameDescriptioncasual口语化Like chatting with friends — natural, relaxedzhihu知乎Rational, in-depth, personal opinionsxiaohongshu小红书Enthusiastic, emoji-rich, product-focusedwechat公众号Storytelling, engaging, relatableacademic学术Rigorous, precise, no colloquialismsliterary文艺Poetic, imagery-rich, metaphoricalweibo微博Short, opinionated, shareable"
      },
      {
        "title": "Combine humanize + style",
        "body": "python scripts/humanize_cn.py text.txt --style xiaohongshu -o xhs.txt\n\nThis first humanizes (removes AI patterns) then applies the style transform."
      },
      {
        "title": "External Configuration",
        "body": "All patterns, replacements, and scoring weights are in scripts/patterns_cn.json. Edit this file to:\n\nAdd new AI vocabulary patterns\nCustomize replacement alternatives\nAdjust scoring weights per severity\nAdd regex patterns for template detection\nSet thresholds for hedging language detection"
      },
      {
        "title": "detect_cn.py",
        "body": "python scripts/detect_cn.py [file] [-j] [-s] [-v] [--sentences N]\n\nFlagDescription-jJSON output-sScore only (e.g. \"72/100 (high)\")-vVerbose: show worst sentences--sentences NNumber of worst sentences to show (default: 5)"
      },
      {
        "title": "humanize_cn.py",
        "body": "python scripts/humanize_cn.py [file] [-o output] [--scene S] [--style S] [-a] [--seed N]\n\nFlagDescription-oOutput file--scenegeneral/social/tech/formal/chat--stylecasual/zhihu/xiaohongshu/wechat/academic/literary/weibo-aAggressive mode--seedRandom seed for reproducibility"
      },
      {
        "title": "style_cn.py",
        "body": "python scripts/style_cn.py [file] --style S [-o output] [--seed N] [--list]"
      },
      {
        "title": "compare_cn.py",
        "body": "python scripts/compare_cn.py [file] [-o output] [--scene S] [--style S] [-a]\n\nShows score diff, category changes, and metric comparison before/after humanization."
      },
      {
        "title": "Workflow",
        "body": "# 1. Check AI score\npython scripts/detect_cn.py document.txt -v\n\n# 2. Humanize with comparison\npython scripts/compare_cn.py document.txt --scene tech -a -o clean.txt\n\n# 3. Verify improvement\npython scripts/detect_cn.py clean.txt -s\n\n# 4. Optional: apply specific style\npython scripts/style_cn.py clean.txt --style zhihu -o final.txt"
      },
      {
        "title": "Batch Processing",
        "body": "# Scan all files\nfor f in *.txt; do\n  echo \"=== $f ===\"\n  python scripts/detect_cn.py \"$f\" -s\ndone\n\n# Transform all markdown\nfor f in *.md; do\n  python scripts/humanize_cn.py \"$f\" --scene tech -a -o \"${f%.md}_clean.md\"\ndone"
      }
    ],
    "body": "Humanize Chinese AI Text v2.0\n\nComprehensive CLI for detecting and transforming Chinese AI-generated text. Makes robotic AI writing natural and human-like.\n\nv2.0 highlights: weighted 0-100 scoring, sentence-level analysis, sentence restructuring (merge/split), context-aware replacement, rhythm variation, vocabulary diversification, 7 style transforms, external pattern config (patterns_cn.json).\n\nQuick Start\n# Detect AI patterns (20+ categories, 0-100 score)\npython scripts/detect_cn.py text.txt\npython scripts/detect_cn.py text.txt -v          # verbose + worst sentences\npython scripts/detect_cn.py text.txt -s           # score only\npython scripts/detect_cn.py text.txt -j           # JSON output\n\n# Humanize text\npython scripts/humanize_cn.py text.txt -o clean.txt\npython scripts/humanize_cn.py text.txt --scene social\npython scripts/humanize_cn.py text.txt --scene tech -a   # aggressive mode\npython scripts/humanize_cn.py text.txt --seed 42         # reproducible\n\n# Apply writing styles\npython scripts/style_cn.py text.txt --style zhihu -o zhihu.txt\npython scripts/style_cn.py text.txt --style xiaohongshu\npython scripts/style_cn.py --list\n\n# Compare before/after\npython scripts/compare_cn.py text.txt --scene tech -a\npython scripts/compare_cn.py text.txt -o clean.txt\n\nDetection System\nScoring\n\nWeighted 0-100 score with 4 severity levels:\n\nScore\tLevel\tMeaning\n0-24\tLOW\tLikely human-written\n25-49\tMEDIUM\tSome AI signals\n50-74\tHIGH\tProbably AI-generated\n75-100\tVERY HIGH\tAlmost certainly AI\nDetection Categories\n🔴 Critical (weight: 8)\nCategory\tExamples\nThree-Part Structure\t首先...其次...最后, 一方面...另一方面, 其一...其二...其三\nMechanical Connectors\t值得注意的是, 综上所述, 不难发现, 归根结底, 由此可见\nEmpty Grand Words\t赋能, 闭环, 数字化转型, 协同增效, 全方位, 多维度\n🟠 High Signal (weight: 4)\nCategory\tExamples\nAI High-Frequency Words\t助力, 彰显, 底层逻辑, 抓手, 触达, 沉淀, 复盘\nFiller Phrases\t值得一提的是, 众所周知, 毫无疑问\nBalanced Arguments\t虽然...但是...同时, 既有...也有...更有\nTemplate Sentences\t随着...的不断发展, 在当今...时代, 作为...的重要组成部分\n🟡 Medium Signal (weight: 2)\nCategory\tExamples\nHedging Language\t在一定程度上, 某种程度上, 通常情况下 (>5 occurrences)\nList Addiction\tExcessive numbered/bulleted lists\nPunctuation Overuse\tDense em dashes, semicolons\nExcessive Rhetoric\t对偶/排比句过多\n⚪ Style Signal (weight: 1.5)\nCategory\tDescription\nUniform Paragraphs\tLow CV in paragraph lengths\nLow Burstiness\tMonotonous sentence lengths\nEmotional Flatness\tLack of emotional/personal expressions\nRepetitive Starters\tSame sentence starters >3 times\nLow Entropy\tLow character-level entropy (predictable text)\nSentence-Level Analysis\n\nWith -v (verbose) mode, the detector identifies the most AI-like sentences:\n\n── 最可疑句子 ──\n  1. [16分] 随着人工智能技术的不断发展，在当今数字化转型时代...\n     原因: 数字化转型, 深度融合, 模板: 随着.*?的(不断)?发展\n\nHumanization Engine\nTransforms (applied in order)\nStructure cleanup — Remove three-part structure (首先/其次/最后)\nPhrase replacement — Context-aware replacement of AI phrases (regex patterns first, then plain text, longest-first matching)\nSentence merge — Merge overly short consecutive sentences\nSentence split — Split long sentences at natural breakpoints (但是/不过/同时)\nPunctuation normalization — Reduce excessive semicolons, em dashes\nVocabulary diversification — Replace repeated words (进行/实现/提供 etc.) with synonyms\nParagraph rhythm — Vary uniform paragraph lengths (merge short, split long)\nCasual injection — Add human expressions (scene-dependent)\nParagraph shortening — For social/chat scenes\nScenes\nScene\tCasualness\tBest For\ngeneral\t0.3\tDefault, balanced\nsocial\t0.7\tSocial media, short posts\ntech\t0.3\tTech blogs, tutorials\nformal\t0.1\tFormal articles, reports\nchat\t0.8\tConversations, messaging\nAggressive Mode (-a)\n\nAdds +0.3 casualness, more colloquial expressions, stronger sentence restructuring. Typical score reduction: 60-80 points on heavily AI-generated text.\n\nReproducibility\n\nUse --seed N for reproducible results (same input + seed = same output).\n\nWriting Style Transforms\n\n7 specialized Chinese writing styles:\n\nStyle\tName\tDescription\ncasual\t口语化\tLike chatting with friends — natural, relaxed\nzhihu\t知乎\tRational, in-depth, personal opinions\nxiaohongshu\t小红书\tEnthusiastic, emoji-rich, product-focused\nwechat\t公众号\tStorytelling, engaging, relatable\nacademic\t学术\tRigorous, precise, no colloquialisms\nliterary\t文艺\tPoetic, imagery-rich, metaphorical\nweibo\t微博\tShort, opinionated, shareable\nCombine humanize + style\npython scripts/humanize_cn.py text.txt --style xiaohongshu -o xhs.txt\n\n\nThis first humanizes (removes AI patterns) then applies the style transform.\n\nExternal Configuration\n\nAll patterns, replacements, and scoring weights are in scripts/patterns_cn.json. Edit this file to:\n\nAdd new AI vocabulary patterns\nCustomize replacement alternatives\nAdjust scoring weights per severity\nAdd regex patterns for template detection\nSet thresholds for hedging language detection\nScripts Reference\ndetect_cn.py\npython scripts/detect_cn.py [file] [-j] [-s] [-v] [--sentences N]\n\nFlag\tDescription\n-j\tJSON output\n-s\tScore only (e.g. \"72/100 (high)\")\n-v\tVerbose: show worst sentences\n--sentences N\tNumber of worst sentences to show (default: 5)\nhumanize_cn.py\npython scripts/humanize_cn.py [file] [-o output] [--scene S] [--style S] [-a] [--seed N]\n\nFlag\tDescription\n-o\tOutput file\n--scene\tgeneral/social/tech/formal/chat\n--style\tcasual/zhihu/xiaohongshu/wechat/academic/literary/weibo\n-a\tAggressive mode\n--seed\tRandom seed for reproducibility\nstyle_cn.py\npython scripts/style_cn.py [file] --style S [-o output] [--seed N] [--list]\n\ncompare_cn.py\npython scripts/compare_cn.py [file] [-o output] [--scene S] [--style S] [-a]\n\n\nShows score diff, category changes, and metric comparison before/after humanization.\n\nWorkflow\n# 1. Check AI score\npython scripts/detect_cn.py document.txt -v\n\n# 2. Humanize with comparison\npython scripts/compare_cn.py document.txt --scene tech -a -o clean.txt\n\n# 3. Verify improvement\npython scripts/detect_cn.py clean.txt -s\n\n# 4. Optional: apply specific style\npython scripts/style_cn.py clean.txt --style zhihu -o final.txt\n\nBatch Processing\n# Scan all files\nfor f in *.txt; do\n  echo \"=== $f ===\"\n  python scripts/detect_cn.py \"$f\" -s\ndone\n\n# Transform all markdown\nfor f in *.md; do\n  python scripts/humanize_cn.py \"$f\" --scene tech -a -o \"${f%.md}_clean.md\"\ndone"
  },
  "trust": {
    "sourceLabel": "tencent",
    "provenanceUrl": "https://clawhub.ai/swaylq/humanize-chinese",
    "publisherUrl": "https://clawhub.ai/swaylq/humanize-chinese",
    "owner": "swaylq",
    "version": "2.0.0",
    "license": null,
    "verificationStatus": "Indexed source record"
  },
  "links": {
    "detailUrl": "https://openagent3.xyz/skills/humanize-chinese",
    "downloadUrl": "https://openagent3.xyz/downloads/humanize-chinese",
    "agentUrl": "https://openagent3.xyz/skills/humanize-chinese/agent",
    "manifestUrl": "https://openagent3.xyz/skills/humanize-chinese/agent.json",
    "briefUrl": "https://openagent3.xyz/skills/humanize-chinese/agent.md"
  }
}