[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"extension-plugin-MohamedAbdallah-14-unslop-zh-CN":3,"guides-for-MohamedAbdallah-14-unslop":619,"similar-k174v4m2d0ncx0vw8gs57bn98n86nh9z-zh-CN":620},{"_creationTime":4,"_id":5,"children":6,"community":146,"display":147,"evaluation":150,"identity":383,"isFallback":369,"parentExtension":386,"providers":414,"relations":420,"repo":421,"tags":616,"workflow":617},1778693887244.665,"k174v4m2d0ncx0vw8gs57bn98n86nh9z",[7,49,75,99,122],{"_creationTime":8,"_id":9,"community":10,"display":12,"identity":18,"providers":24,"relations":40,"tags":44,"workflow":45},1778693888107.5203,"k17dxk12pc7h8bstrrbwz0h1bd86mnf9",{"reviewCount":11},0,{"description":13,"installMethods":14,"name":16,"sourceUrl":17},"使 LLM 的输出人性化，使其读起来像一个谨慎的人写的那样。减去 AI 常用语（谄媚、三段式、过多使用破折号、“深入探究”/“挂毯”/“证明”等词、套话、整齐的五段式结构），调整其爆发力和可控的不确定性，并保留技术准确性。支持强度级别：细微、平衡（默认）、完全、声音匹配、反检测器。当用户说“人性化”、“使其听起来像人”、“去粗俗化”、“不带 AI 语调重写”、“匹配我的声音”、“不那么机器人化”或调用 /unslop 时使用。当请求文本质量时也会自动触发。\n",{"claudeCode":15},"MohamedAbdallah-14/unslop","Unslop","https://github.com/MohamedAbdallah-14/unslop",{"basePath":19,"githubOwner":20,"githubRepo":21,"locale":22,"slug":21,"type":23},"skills/unslop","MohamedAbdallah-14","unslop","zh-CN","skill",{"evaluate":25,"extract":37},{"promptVersionExtension":26,"promptVersionScoring":27,"score":28,"tags":29,"targetMarket":35,"tier":36},"3.0.0","4.4.0",98,[30,31,32,33,34],"markdown","humanize","ai-writing","text-processing","cliche-removal","global","verified",{"commitSha":38,"license":39},"HEAD","MIT",{"parentExtensionId":41,"repoId":42,"translatedFrom":43},"k177fsagh49r77m9y4755zc1mn86m1jm","kd727xcarpnqcat3wd68ms466s86mwkb","k17a0sk2mmgjvcn0hqj8tqvbx986nzz5",[32,34,31,30,33],{"evaluatedAt":46,"extractAt":47,"updatedAt":48},1778693789329,1778693661691,1778693888107,{"_creationTime":50,"_id":51,"community":52,"display":53,"identity":57,"providers":60,"relations":69,"tags":71,"workflow":72},1778693894103.9775,"k1797h1np600c3xdf6q3n76db586mgmn",{"reviewCount":11},{"description":54,"installMethods":55,"name":56,"sourceUrl":17},"重写提交消息，使其听起来像一个严谨的程序员编写的。去除 AI/营销术语（“全面的解决方案”、“健壮的实现”、“利用”、“增强”、“无缝地”、“此提交...”）。保留 Conventional Commits 格式。主题 ≤72 个字符（目标 ≤50），祈使语气。仅当“为什么”在主题中不明显时才写正文。在用户说“人性化提交”、“去营销术语化提交消息”、“让这个提交听起来像人话”、“/unslop-commit”、“/commit”、“写一个提交”或粘贴草稿提交以供清理时使用。暂存更改时自动触发。\n",{"claudeCode":15},"Unslop Commit",{"basePath":58,"githubOwner":20,"githubRepo":21,"locale":22,"slug":59,"type":23},"skills/unslop-commit","unslop-commit",{"evaluate":61,"extract":68},{"promptVersionExtension":26,"promptVersionScoring":27,"score":62,"tags":63,"targetMarket":35,"tier":36},100,[64,65,66,32,67],"commit-message","conventional-commits","git","natural-language-processing",{"commitSha":38,"license":39},{"parentExtensionId":41,"repoId":42,"translatedFrom":70},"k178pmb87re7z9as1b8ky4h9m586ncf7",[32,64,65,66,67],{"evaluatedAt":73,"extractAt":47,"updatedAt":74},1778693755963,1778693894104,{"_creationTime":76,"_id":77,"community":78,"display":79,"identity":83,"providers":85,"relations":93,"tags":95,"workflow":96},1778693903062.1482,"k172escz80brxd36dc0zarxwas86n87z",{"reviewCount":11},{"description":80,"installMethods":81,"name":82,"sourceUrl":17},"unslop 模式、子技能和斜杠命令的快速参考卡。一次性显示，不是持久模式。触发方式：/unslop-help、\"unslop help\"、\"what unslop commands\"、\"how do I use unslop\"。\n",{"claudeCode":15},"unslop-help",{"basePath":84,"githubOwner":20,"githubRepo":21,"locale":22,"slug":82,"type":23},"skills/unslop-help",{"evaluate":86,"extract":92},{"promptVersionExtension":26,"promptVersionScoring":27,"score":87,"tags":88,"targetMarket":35,"tier":36},99,[32,33,89,90,91],"llm-utility","prompt-engineering","developer-tool",{"commitSha":38},{"parentExtensionId":41,"repoId":42,"translatedFrom":94},"k171cmam2qfe7534nk47mycyk186nhkp",[32,91,89,90,33],{"evaluatedAt":97,"extractAt":47,"updatedAt":98},1778693807425,1778693903062,{"_creationTime":100,"_id":101,"community":102,"display":103,"identity":107,"providers":109,"relations":116,"tags":118,"workflow":119},1778693903390.1921,"k17d7h61hjw5ay2pewvm72870n86nsey",{"reviewCount":11},{"description":104,"installMethods":105,"name":106,"sourceUrl":17},"从推理过程（思维链、扩展思考、代理分解）中剥离 AI 噪音模式——非最终文案。推理文本有其自身的噪音目录，常规 unslop 无法针对：过度解释问题、过度设限、将微不足道的问题分解为 6 个要点的子步骤、无限循环的合理化。触发方式：/unslop-reasoning，\"清理我的推理\"，\"修复这个思维链\"，\"这个 CoT 听起来很机械\"。适用于推理输出；不覆盖常规 /unslop 模式。\n",{"claudeCode":15},"unslop-reasoning",{"basePath":108,"githubOwner":20,"githubRepo":21,"locale":22,"slug":106,"type":23},"skills/unslop-reasoning",{"evaluate":110,"extract":115},{"promptVersionExtension":26,"promptVersionScoring":27,"score":28,"tags":111,"targetMarket":35,"tier":36},[32,112,113,67,114],"prose","editing","llm-tools",{"commitSha":38},{"parentExtensionId":41,"repoId":42,"translatedFrom":117},"k17652fqyq6b3erz6edk1syfm986m91c",[32,113,114,67,112],{"evaluatedAt":120,"extractAt":47,"updatedAt":121},1778693825172,1778693903390,{"_creationTime":123,"_id":124,"community":125,"display":126,"identity":130,"providers":132,"relations":140,"tags":142,"workflow":143},1778693920987.904,"k173acn013r9q2bkahde5zsrpd86nc7j",{"reviewCount":11},{"description":127,"installMethods":128,"name":129,"sourceUrl":17},"重写代码审查意见，使其听起来像人类同事所写。消除 \"I noticed...\", \"I was wondering if perhaps...\", \"It might be worth considering...\" 这类企业AI的套话。每条评论都直截了当：地点、问题、具体修复方案。用于用户说“人性化审查”、“减少PR评论中的废话”、“让反馈听起来更人性化”、“审查此PR”、“代码审查”、“/review”、“/unslop-review”时。在审查PR时自动触发。\n",{"claudeCode":15},"unslop-review",{"basePath":131,"githubOwner":20,"githubRepo":21,"locale":22,"slug":129,"type":23},"skills/unslop-review",{"evaluate":133,"extract":139},{"promptVersionExtension":26,"promptVersionScoring":27,"score":62,"tags":134,"targetMarket":35,"tier":36},[135,136,137,67,138],"text-rewriting","ai-detection","code-review","developer-tools",{"commitSha":38},{"parentExtensionId":41,"repoId":42,"translatedFrom":141},"k176ahvt49a4bskq406thpyn6x86n6k4",[136,137,138,67,135],{"evaluatedAt":144,"extractAt":47,"updatedAt":145},1778693837524,1778693920987,{"reviewCount":11},{"description":148,"installMethods":149,"name":21,"sourceUrl":17},"使助手输出听起来更人性化。去除 AI 术语（谄媚、陈词滥调、敷衍的说法、连用的破折号），营造自然的爆发力，恢复语音。保留代码、URL 和技术准确性。",{"claudeCode":21},{"_creationTime":151,"_id":152,"extensionId":5,"locale":22,"result":153,"trustSignals":367,"workflow":381},1778693887244.6653,"kn7ctdth8665q5h8mw129rrc0h86n84c",{"checks":154,"evaluatedAt":339,"extensionSummary":340,"features":341,"nonGoals":347,"promptVersionExtension":26,"promptVersionScoring":27,"purpose":351,"rationale":352,"score":62,"summary":353,"tags":354,"tier":36,"useCases":362},[155,160,163,166,170,173,177,181,184,187,191,195,198,202,205,208,211,214,217,220,224,228,232,236,240,243,246,249,253,256,259,262,265,268,271,275,279,283,287,291,294,297,300,303,306,309,312,315,318,321,325,328,331,335],{"category":156,"check":157,"severity":158,"summary":159},"Practical Utility","Problem relevance","pass","描述清楚地阐述了 AI 生成的文本听起来不自然的问题，并提出了使其听起来更人性化的解决方案。",{"category":156,"check":161,"severity":158,"summary":162},"Unique selling proposition","该扩展提供了独特的卖点，专注于去除 AI 术语同时保留事实内容和技术准确性，这是一种超越基本 LLM 提示的特定价值。",{"category":156,"check":164,"severity":158,"summary":165},"Production readiness","该插件已准备好投入生产，支持多种 IDE 和 CLI 环境，提供多种模式，并提供清晰的安装和使用说明，涵盖了其声明的用例。",{"category":167,"check":168,"severity":158,"summary":169},"Scope","Single responsibility principle","该插件专注于人性化 AI 生成文本的单一领域，其子功能是与此主要目的保持一致的扩展。",{"category":167,"check":171,"severity":158,"summary":172},"Description quality","显示的描述准确简洁地反映了插件的核心功能和价值主张。",{"category":174,"check":175,"severity":158,"summary":176},"Invocation","Scoped tools","该插件公开了狭窄的、动词-名词专家的工具，如 `/unslop \u003Cmode>`、`/unslop-commit` 和 `/unslop-review`，避免了广泛的、自由格式的命令执行。",{"category":178,"check":179,"severity":158,"summary":180},"Documentation","Configuration & parameter reference","文档清楚地解释了所有模式、子功能和配置选项，包括默认模式的持久化和解析顺序。",{"category":167,"check":182,"severity":158,"summary":183},"Tool naming","所有公开的工具名称（/unslop、/unslop-commit 等）都具有描述性，采用烤串命名法，并且清晰地与插件的领域相关。",{"category":167,"check":185,"severity":158,"summary":186},"Minimal I/O surface","工具通过特定的命令或自然语言短语触发，输出集中在重写后的文本或状态消息上，而不是内部诊断。",{"category":188,"check":189,"severity":158,"summary":190},"License","License usability","该扩展根据宽松的 MIT 许可证分发，通过 LICENSE 文件和 README 中的 shield 明确标示。",{"category":192,"check":193,"severity":158,"summary":194},"Maintenance","Commit recency","上次提交是在 2026 年 5 月 5 日，远在 3 个月阈值之内，表明维护活跃。",{"category":192,"check":196,"severity":158,"summary":197},"Dependency Management","该插件似乎具有最少的外部依赖项，并且 README 没有指出其管理方面的问题。Node.js 被提及为 hooks 的要求。",{"category":199,"check":200,"severity":158,"summary":201},"Security","Secret Management","该插件似乎不直接处理密钥；它依赖于主机环境的 API 密钥，并且不公开它们。",{"category":199,"check":203,"severity":158,"summary":204},"Injection","该插件旨在处理文本和代码，并且 README 强调保留代码和技术准确性，这表明它不会从外部数据中遵循指令。",{"category":199,"check":206,"severity":158,"summary":207},"Transitive Supply-Chain Grenades","该插件的 hooks 是仓库内的 JavaScript 和 Bash/PowerShell 脚本，CLI 是 Python 包。没有运行时下载外部代码或数据的迹象。",{"category":199,"check":209,"severity":158,"summary":210},"Sandbox Isolation","该插件在主机环境（Claude Code、CLI）中本地运行，并且其 hooks 设计用于修改 `settings.json` 或发出文本，而不是触碰外部文件或路径。",{"category":199,"check":212,"severity":158,"summary":213},"Sandbox escape primitives","hook 脚本是用纯 JavaScript 和 Bash/PowerShell 编写的，没有显示分离的进程或在拒绝的工具调用周围进行重试循环的迹象。",{"category":199,"check":215,"severity":158,"summary":216},"Data Exfiltration","README 明确说明“没有遥测、没有分析、没有回传”，并且该插件在本地运行或使用用户提供的 API 密钥用于主机环境。",{"category":199,"check":218,"severity":158,"summary":219},"Hidden Text Tricks","捆绑的内容和描述似乎没有隐藏的控制技巧、不可见的 Unicode 字符或其他混淆方法。",{"category":221,"check":222,"severity":158,"summary":223},"Hooks","Opaque code execution","hook 脚本是纯 JavaScript，CLI 是 Python。没有混淆的代码、最小化的捆绑包或运行时 `eval` 模式。",{"category":225,"check":226,"severity":158,"summary":227},"Portability","Structural Assumption","该插件将 hooks 安装到标准的 Claude/用户配置目录中，CLI 通过 pip 安装，避免了对用户项目结构的假设。",{"category":229,"check":230,"severity":158,"summary":231},"Trust","Issues Attention","在过去 90 天内，0 个打开和 0 个关闭的 issue，该插件似乎是新近的或稳定的，没有当前问题，符合通过标准。",{"category":233,"check":234,"severity":158,"summary":235},"Versioning","Release Management","该插件在其安装命令中使用语义化版本控制，并有一个更改日志，表明版本控制实践清晰。",{"category":237,"check":238,"severity":158,"summary":239},"Code Execution","Validation","hooks 和 CLI 脚本似乎通过验证和安全写入实践来处理输入模式和文件路径，如配置和安装程序逻辑所示。",{"category":199,"check":241,"severity":158,"summary":242},"Unguarded Destructive Operations","该插件专注于文本操作和代码注入，而不是破坏性操作。文件重写工具会备份原始文件，从而降低了风险。",{"category":237,"check":244,"severity":158,"summary":245},"Error Handling","hook 和 CLI 脚本采用了 try/catch 块、安全的文件操作以及对有效模式和配置的显式检查，表明了健壮的错误处理。",{"category":237,"check":247,"severity":158,"summary":248},"Logging","该插件明确说明“没有遥测、没有分析、没有回传”。未提及本地审计日志记录，但对于此类实用程序并非必需。",{"category":250,"check":251,"severity":158,"summary":252},"Compliance","GDPR","该插件不处理个人数据；它在本地处理文本和代码文件，并且不将数据传输给第三方。",{"category":250,"check":254,"severity":158,"summary":255},"Target market","该扩展专为代码编辑环境和自然语言处理而设计，没有任何明显的区域或司法管辖区限制；targetMarket 为 'global'。",{"category":225,"check":257,"severity":158,"summary":258},"Runtime stability","hooks 用 Node.js 和 Bash/PowerShell 编写，与标准环境兼容。Python CLI 依赖于标准库和 pip，可确保跨平台兼容性。",{"category":178,"check":260,"severity":158,"summary":261},"README","提供了全面的 README，清晰地说明了扩展的目的、功能、安装和用法，并附有示例。",{"category":167,"check":263,"severity":158,"summary":264},"Tool surface size","该插件暴露了合理数量的工具（5 个主命令 + 模式），处于 3-10 的目标范围内。",{"category":174,"check":266,"severity":158,"summary":267},"Overlapping near-synonym tools","工具具有不同的名称和目的，并且每个功能都有清晰的触发器（例如，/unslop 用于常规模式，/unslop-commit 用于提交）。",{"category":178,"check":269,"severity":158,"summary":270},"Phantom features","README 中描述的所有功能，包括各种模式和子功能，都有相应的实现和文档。",{"category":272,"check":273,"severity":158,"summary":274},"Install","Installation instruction","为 Claude Code、CLI 和其他环境提供了清晰、可复制粘贴的安装说明，包括 hooks 安装。",{"category":276,"check":277,"severity":158,"summary":278},"Errors","Actionable error messages","脚本和 hook 逻辑包括特定的错误消息和回退机制，指导用户了解潜在问题。",{"category":280,"check":281,"severity":158,"summary":282},"Execution","Pinned dependencies","通过 pip 进行的 Python 包安装意味着依赖项管理，hook 脚本是纯 JS/Bash/PS，除了 Node.js 之外没有外部运行时依赖项。",{"category":167,"check":284,"severity":285,"summary":286},"Dry-run preview","not_applicable","该扩展的核心功能是文本重写和样式调整，不涉及需要干运行模式的状态更改操作或出站数据发送。",{"category":288,"check":289,"severity":158,"summary":290},"Protocol","Idempotent retry & timeouts","hooks 具有显式超时，CLI 操作主要是文本转换或本地文件操作，不需要复杂的重试或有状态逻辑。",{"category":250,"check":292,"severity":158,"summary":293},"Telemetry opt-in","该插件明确表示不收集遥测或分析数据，默认情况下满足选择加入的要求（关闭）。",{"category":174,"check":295,"severity":158,"summary":296},"Name collisions","该插件的命令（/unslop、/unslop-commit 等）是不同的，并且似乎没有与 Claude Code 内置命令或其他常见命令发生冲突。",{"category":174,"check":298,"severity":158,"summary":299},"Hooks-off mechanism","该插件提供了清晰的“停止 unslop”/“正常模式”命令，并且可以通过 `UNSLOP_DEFAULT_MODE=off` 或删除标志文件来禁用自动激活。",{"category":174,"check":301,"severity":158,"summary":302},"Hook matcher tightness","hooks 由特定命令（`/unslop`）、自然语言模式或会话开始触发，而不是广泛的通配符。",{"category":199,"check":304,"severity":158,"summary":305},"Hook security","hooks 主要修改文本或管理本地状态标志；它们不执行破坏性操作或在没有用户调用的情况下进行网络调用。",{"category":221,"check":307,"severity":158,"summary":308},"Silent prompt rewriting","该插件的 hooks 修改输出或上下文，而不是直接修改用户的提示，并且任何状态更新都将在 UI 中可见。",{"category":199,"check":310,"severity":285,"summary":311},"Permission Hook","该插件不与 `PermissionRequest` hook 交互。",{"category":250,"check":313,"severity":158,"summary":314},"Hook privacy","该插件明确避免遥测和数据传输，hooks 在本地处理标志或文本转换。",{"category":237,"check":316,"severity":158,"summary":317},"Hook dependency","hook 脚本是仓库中简短、可读的 JavaScript/Bash/PowerShell 文件，而不是不透明的捆绑包或运行时获取。",{"category":178,"check":319,"severity":158,"summary":320},"Feature Transparency","所有声明的 hooks 及其功能都在 README 和相关的 SKILL.md 文件中进行了说明。",{"category":322,"check":323,"severity":158,"summary":324},"Convention","Layout convention adherence","该插件遵循 Claude Code 的约定，在 `.claude-plugin/plugin.json` 中包含插件信息，并在 `hooks/` 目录中放置 hooks，并且 bin 条目已文档化。",{"category":322,"check":326,"severity":158,"summary":327},"Plugin state","插件状态（如活动模式标志）在 `$CLAUDE_CONFIG_DIR` 下管理，确保它由默认卸载语义处理。",{"category":199,"check":329,"severity":285,"summary":330},"Keychain-stored secrets","该插件不通过 `userConfig` 或其他方式消耗密钥；它依赖于主机环境的身份验证。",{"category":332,"check":333,"severity":158,"summary":334},"Dependencies","Tagged release sourcing","该插件直接通过 marketplace 从 GitHub 存储库安装，其自身的依赖项（Node.js、Python）是标准的。",{"category":336,"check":337,"severity":158,"summary":338},"Installation","Clean uninstall","提供的卸载脚本会删除 hook 文件并清理 settings.json 条目，确保干净卸载而没有后台守护进程。",1778693722536,"该插件提供多种模式来重写 AI 生成的文本，通过去除常见的 AI 模式（如谄媚、陈词滥调和敷衍的说法）使其听起来更人性化。它保留代码、URL 和技术准确性，支持多种 IDE 和命令行界面。",[342,343,344,345,346],"去除 AI 术语（谄媚、陈词滥调、敷衍的说法）","营造自然的爆发力和节奏感","保留代码、URL 和技术准确性","提供多种模式（微妙、平衡、完整、声音匹配、反检测器）","支持与 Claude Code、Cursor、Windsurf 和 CLI 集成",[348,349,350],"声称使文本 100% 无法被 AI 检测器检测","用风格偏好取代事实准确性","进行学术不端行为或抄袭","使 AI 生成的文本听起来更自然、更像人类，从而提高其可读性并在需要个人声音的场景中提高接受度。","所有检查均通过。该插件文档齐全、安全且具有明确的实用性。","一个高质量的插件，通过去除“AI 术语”和提高可读性来增强 AI 生成的文本。",[355,356,357,358,359,360,361],"ai","text","writing","editor","code","nlp","humanizer",[363,364,365,366],"使求职信和简历更具人性化","改善营销文案的语气","让 AI 生成的博客文章更具吸引力","优化 AI 生成的代码审查评论",{"codeQuality":368,"collectedAt":370,"documentation":371,"maintenance":374,"security":379,"testCoverage":380},{"hasLockfile":369},false,1778693702825,{"descriptionLength":372,"readmeSize":373},190,52819,{"closedIssues90d":11,"forks":375,"hasChangelog":376,"openIssues90d":11,"pushedAt":377,"stars":378},1,true,1777948198000,36,{"hasNpmPackage":369,"license":39,"smitheryVerified":369},{"hasCi":376,"hasTests":376},{"updatedAt":382},1778693887244,{"basePath":384,"githubOwner":20,"githubRepo":21,"locale":22,"slug":21,"type":385},"","plugin",{"_creationTime":387,"_id":388,"community":389,"display":390,"identity":394,"parentExtension":397,"providers":398,"relations":409,"tags":410,"workflow":411},1778693661691.4355,"k175vxsqnmn2ye2xkw62x4enkh86n8eb",{"reviewCount":11},{"description":391,"installMethods":392,"name":393,"sourceUrl":17},"Plugins that make model-assisted text sound natural and human: clearer voice, less robotic phrasing, better burstiness, no AI fingerprint.",{"claudeCode":15},"unslop-marketplace",{"basePath":384,"githubOwner":20,"githubRepo":21,"locale":395,"slug":21,"type":396},"en","marketplace",null,{"evaluate":399,"extract":405},{"promptVersionExtension":400,"promptVersionScoring":27,"score":62,"tags":401,"targetMarket":35,"tier":36},"3.1.0",[357,402,136,403,404],"humanization","voice","natural-language",{"commitSha":38,"marketplace":406,"plugin":407},{"name":393,"pluginCount":375},{"mcpCount":11,"provider":408,"skillCount":11},"classify",{"repoId":42},[136,402,404,403,357],{"evaluatedAt":412,"extractAt":47,"updatedAt":413},1778693702554,1778693923493,{"evaluate":415,"extract":417},{"promptVersionExtension":26,"promptVersionScoring":27,"score":62,"tags":416,"targetMarket":35,"tier":36},[355,356,357,358,359,360,361],{"commitSha":38,"plugin":418},{"mcpCount":11,"provider":408,"skillCount":419},5,{"parentExtensionId":388,"repoId":42,"translatedFrom":41},{"_creationTime":422,"_id":42,"identity":423,"providers":424,"workflow":612},1778693654114.4695,{"githubOwner":20,"githubRepo":21,"sourceUrl":17},{"classify":425,"discover":593,"github":596},{"commitSha":38,"extensions":426},[427,438,449,505,512,518,565,571,577,583],{"basePath":428,"description":429,"displayName":430,"installMethods":431,"rationale":432,"selectedPaths":433,"source":437,"sourceLanguage":395,"type":396},".agents/plugins","Local agents marketplace for the unslop plugin (anti-AI-slop tooling).","unslop-agents-marketplace",{"claudeCode":15},"marketplace.json at .agents/plugins/marketplace.json",[434],{"path":435,"priority":436},"marketplace.json","mandatory","rule",{"basePath":384,"description":391,"displayName":393,"installMethods":439,"rationale":440,"selectedPaths":441,"source":437,"sourceLanguage":395,"type":396},{"claudeCode":15},"marketplace.json at .claude-plugin/marketplace.json",[442,444,446],{"path":443,"priority":436},".claude-plugin/marketplace.json",{"path":445,"priority":436},"README.md",{"path":447,"priority":448},"LICENSE","high",{"basePath":384,"description":450,"displayName":21,"installMethods":451,"rationale":452,"selectedPaths":453,"source":437,"sourceLanguage":395,"type":385},"Make assistant output sound human. Strip AI-isms (sycophancy, stock vocab, hedging stacks, em-dash pileups), engineer burstiness, restore voice. Preserves code, URLs, and technical accuracy.",{"claudeCode":21},"plugin manifest at .claude-plugin/plugin.json (coalesced with duplicate plugin at plugins/unslop)",[454,456,457,458,461,463,465,467,469,471,473,475,477,479,481,483,485,487,489,491,493,495,497,499,501,503],{"path":455,"priority":436},".claude-plugin/plugin.json",{"path":445,"priority":436},{"path":447,"priority":448},{"path":459,"priority":460},"skills/unslop/SKILL.md","medium",{"path":462,"priority":460},"skills/unslop-commit/SKILL.md",{"path":464,"priority":460},"skills/unslop-file/SKILL.md",{"path":466,"priority":460},"skills/unslop-help/SKILL.md",{"path":468,"priority":460},"skills/unslop-reasoning/SKILL.md",{"path":470,"priority":460},"skills/unslop-review/SKILL.md",{"path":472,"priority":448},"commands/unslop-commit.toml",{"path":474,"priority":448},"commands/unslop-file.toml",{"path":476,"priority":448},"commands/unslop-help.toml",{"path":478,"priority":448},"commands/unslop-reasoning.toml",{"path":480,"priority":448},"commands/unslop-review.toml",{"path":482,"priority":448},"commands/unslop.toml",{"path":484,"priority":448},"hooks/README.md",{"path":486,"priority":448},"hooks/install.ps1",{"path":488,"priority":448},"hooks/install.sh",{"path":490,"priority":448},"hooks/package.json",{"path":492,"priority":448},"hooks/uninstall.ps1",{"path":494,"priority":448},"hooks/uninstall.sh",{"path":496,"priority":448},"hooks/unslop-activate.js",{"path":498,"priority":448},"hooks/unslop-config.js",{"path":500,"priority":448},"hooks/unslop-mode-tracker.js",{"path":502,"priority":448},"hooks/unslop-statusline.ps1",{"path":504,"priority":448},"hooks/unslop-statusline.sh",{"basePath":19,"description":506,"displayName":21,"installMethods":507,"rationale":508,"selectedPaths":509,"source":437,"sourceLanguage":395,"type":23},"Humanize LLM output so it reads like a careful human wrote it. Subtracts AI-isms (sycophancy, tricolons, em-dash overuse, \"delve\"/\"tapestry\"/\"testament\", hedging stacks, tidy five-paragraph shapes), engineers burstiness and calibrated uncertainty, and preserves technical accuracy. Supports intensity levels: subtle, balanced (default), full, voice-match, anti-detector. Use when user says \"humanize this\", \"make this sound human\", \"de-slop this\", \"rewrite without AI tone\", \"match my voice\", \"less robotic\", or invokes /unslop. Also auto-triggers when text-quality is requested.\n",{"claudeCode":15},"SKILL.md frontmatter at skills/unslop/SKILL.md (coalesced with duplicate skill at .cursor/skills/unslop, .windsurf/skills/unslop, plugins/unslop/skills/unslop)",[510],{"path":511,"priority":436},"SKILL.md",{"basePath":58,"description":513,"displayName":59,"installMethods":514,"rationale":515,"selectedPaths":516,"source":437,"sourceLanguage":395,"type":23},"Rewrites commit messages so they sound like a careful human engineer wrote them. Strips AI/marketing slop (\"comprehensive solution\", \"robust implementation\", \"leverage\", \"enhance\", \"seamlessly\", \"This commit...\"). Keeps Conventional Commits format. Subject ≤72 chars (aim ≤50), imperative mood. Body only when \"why\" isn't obvious from the subject. Use when user says \"humanize commit\", \"de-slop commit message\", \"make this commit sound human\", \"/unslop-commit\", \"/commit\", \"write a commit\", or pastes a draft commit to clean up. Auto-triggers when staging changes.\n",{"claudeCode":15},"SKILL.md frontmatter at skills/unslop-commit/SKILL.md (coalesced with duplicate skill at plugins/unslop/skills/unslop-commit)",[517],{"path":511,"priority":436},{"basePath":21,"description":519,"displayName":520,"installMethods":521,"rationale":522,"selectedPaths":523,"source":437,"sourceLanguage":395,"type":23},"Humanize natural-language memory files (CLAUDE.md, todos, preferences, docs) by removing AI-isms and adding burstiness while preserving every code block, URL, path, command, and heading exactly. Two modes: --deterministic (fast, regex-based, no API) and LLM (default, calls Claude for rewrite). Humanized version overwrites the original file. Plain backup saved as FILE.original.md. Trigger: /unslop-file \u003Cfilepath> or \"humanize memory file\"\n","unslop-file",{"claudeCode":15},"SKILL.md frontmatter at unslop/SKILL.md (coalesced with duplicate skill at skills/unslop-file, plugins/unslop/skills/unslop-file)",[524,525,526,528,530,533,535,537,539,541,543,545,547,549,551,553,555,557,559,561,563],{"path":511,"priority":436},{"path":445,"priority":448},{"path":527,"priority":460},"CHANGELOG.md",{"path":529,"priority":460},"SECURITY.md",{"path":531,"priority":532},"scripts/__init__.py","low",{"path":534,"priority":532},"scripts/__main__.py",{"path":536,"priority":532},"scripts/benchmark.py",{"path":538,"priority":532},"scripts/cli.py",{"path":540,"priority":532},"scripts/detect.py",{"path":542,"priority":532},"scripts/detector.py",{"path":544,"priority":532},"scripts/fetch_detectors.py",{"path":546,"priority":532},"scripts/humanize.py",{"path":548,"priority":532},"scripts/lexical_targets.py",{"path":550,"priority":532},"scripts/py.typed",{"path":552,"priority":532},"scripts/reasoning.py",{"path":554,"priority":532},"scripts/soul.py",{"path":556,"priority":532},"scripts/structural.py",{"path":558,"priority":532},"scripts/style_memory.py",{"path":560,"priority":532},"scripts/stylometry.py",{"path":562,"priority":532},"scripts/surprisal.py",{"path":564,"priority":532},"scripts/validate.py",{"basePath":84,"description":566,"displayName":82,"installMethods":567,"rationale":568,"selectedPaths":569,"source":437,"sourceLanguage":395,"type":23},"Quick-reference card for unslop modes, sub-skills, and slash commands. One-shot display, not a persistent mode. Trigger: /unslop-help, \"unslop help\", \"what unslop commands\", \"how do I use unslop\".\n",{"claudeCode":15},"SKILL.md frontmatter at skills/unslop-help/SKILL.md (coalesced with duplicate skill at plugins/unslop/skills/unslop-help)",[570],{"path":511,"priority":436},{"basePath":108,"description":572,"displayName":106,"installMethods":573,"rationale":574,"selectedPaths":575,"source":437,"sourceLanguage":395,"type":23},"Strip AI-slop patterns from reasoning traces (chain-of-thought, extended thinking, agent decomposition) — not final prose. Reasoning text has its own slop catalog that regular unslop doesn't target: over-explaining the question, over-hedging, over-decomposing trivial problems into 6-bullet substeps, infinite-loop rationalization. Trigger: /unslop-reasoning, \"clean up my reasoning\", \"fix this chain of thought\", \"this CoT sounds robotic\". Applies to reasoning output; does not override regular /unslop mode.\n",{"claudeCode":15},"SKILL.md frontmatter at skills/unslop-reasoning/SKILL.md (coalesced with duplicate skill at plugins/unslop/skills/unslop-reasoning)",[576],{"path":511,"priority":436},{"basePath":131,"description":578,"displayName":129,"installMethods":579,"rationale":580,"selectedPaths":581,"source":437,"sourceLanguage":395,"type":23},"Rewrites code review comments so they read like a human teammate wrote them. Cuts corporate-AI throat-clearing (\"I noticed...\", \"I was wondering if perhaps...\", \"It might be worth considering...\"). Each comment is direct: location, the issue, a concrete fix. Use when user says \"humanize review\", \"de-slop PR comment\", \"make this feedback sound human\", \"review this PR\", \"code review\", \"/review\", \"/unslop-review\". Auto-triggers when reviewing PRs.\n",{"claudeCode":15},"SKILL.md frontmatter at skills/unslop-review/SKILL.md (coalesced with duplicate skill at plugins/unslop/skills/unslop-review)",[582],{"path":511,"priority":436},{"basePath":21,"installMethods":584,"rationale":585,"selectedPaths":586,"source":437,"sourceLanguage":395,"type":592},{"pypi":21},"cli ecosystem detected at unslop",[587,589,590,591],{"path":588,"priority":436},"pyproject.toml",{"path":445,"priority":436},{"path":538,"priority":460},{"path":534,"priority":532},"cli",{"sources":594},[595],"manual",{"closedIssues90d":11,"description":597,"forks":375,"homepage":598,"license":39,"openIssues90d":11,"pushedAt":377,"readmeSize":373,"stars":378,"topics":599},"Make AI output sound human. Strips AI-isms (sycophancy, stock vocab, hedging stacks, em-dash pileups), preserves code/URLs/headings. Plugin for Claude Code, Cursor, Windsurf, Codex, Cline, Copilot, Gemini.","https://mohamedabdallah-14.github.io/unslop/",[136,600,601,602,603,604,605,361,606,90,607,32,608,609,610,611],"ai-plugin","anti-slop","claude-code","claude-code-plugin","cline-rules","cursor-rules","llm","windsurf-rules","content-quality","open-source","typescript","writing-tools",{"classifiedAt":613,"discoverAt":614,"extractAt":615,"githubAt":615,"updatedAt":613},1778693661514,1778693654114,1778693659782,[355,359,358,361,360,356,357],{"evaluatedAt":618,"extractAt":47,"updatedAt":382},1778693722676,[],[621,651,681,715,741,764],{"_creationTime":622,"_id":623,"community":624,"display":625,"identity":631,"providers":635,"relations":644,"tags":647,"workflow":648},1778690773482.4834,"k179sm2kkyd7r7nz9jsx62jm9x86mw4a",{"reviewCount":11},{"description":626,"installMethods":627,"name":629,"sourceUrl":630},"Look up and read Hugging Face paper pages in markdown, and use the papers API for structured metadata like authors, linked models, datasets, Spaces, and media URLs when needed.",{"claudeCode":628},"huggingface-papers","Hugging Face Papers","https://github.com/huggingface/skills",{"basePath":632,"githubOwner":633,"githubRepo":634,"locale":395,"slug":628,"type":385},"skills/huggingface-papers","huggingface","skills",{"evaluate":636,"extract":642},{"promptVersionExtension":26,"promptVersionScoring":27,"score":62,"tags":637,"targetMarket":35,"tier":36},[633,638,639,355,640,641],"papers","arxiv","research","metadata",{"commitSha":38,"license":643},"Apache-2.0",{"parentExtensionId":645,"repoId":646},"k17es3r8wd37t5rrwqcpp5kwrh86mxx8","kd72xwt5xnc0ktc4p7smzfcp3986m959",[355,639,633,641,638,640],{"evaluatedAt":649,"extractAt":650,"updatedAt":649},1778690901306,1778690773482,{"_creationTime":652,"_id":653,"community":654,"display":655,"identity":660,"providers":663,"relations":672,"tags":676,"workflow":677},1778685915634.952,"k178sazsw9mc93tarpmxx6wwf586ncry",{"reviewCount":11},{"description":656,"installMethods":657,"name":658,"sourceUrl":659},"创建、更新和修复 Cypress 测试。连接到 Cypress Cloud 以查看测试结果并利用数据来管理您的测试套件。",{"claudeCode":658},"cypress","https://github.com/cypress-io/ai-toolkit",{"basePath":384,"githubOwner":661,"githubRepo":662,"locale":22,"slug":662,"type":385},"cypress-io","ai-toolkit",{"evaluate":664,"extract":669},{"promptVersionExtension":26,"promptVersionScoring":27,"score":62,"tags":665,"targetMarket":35,"tier":36},[658,666,667,355,668],"testing","automation","qa",{"commitSha":38,"license":39,"plugin":670},{"mcpCount":11,"provider":408,"skillCount":671},3,{"parentExtensionId":673,"repoId":674,"translatedFrom":675},"k170k28hx0d93ds1md7v66h33986nap6","kd778b5hp7aqcpb58zn9yj8xas86meqd","k17a80t18qpe9tmapz3fnw597986mpsy",[355,667,658,668,666],{"evaluatedAt":678,"extractAt":679,"updatedAt":680},1778685834132,1778685765056,1778685915635,{"_creationTime":682,"_id":683,"community":684,"display":685,"identity":691,"providers":694,"relations":708,"tags":711,"workflow":712},1778683100520.2961,"k1754vkdjckrkqvz9x7tjrvhzn86n1gc",{"reviewCount":11},{"description":686,"installMethods":687,"name":689,"sourceUrl":690},"AI music generation workflow for Suno - album concepts, lyrics, prompts, mastering, release",{"claudeCode":688},"bitwize-music","Claude AI Music Skills","https://github.com/bitwize-music-studio/claude-ai-music-skills",{"basePath":384,"githubOwner":692,"githubRepo":693,"locale":395,"slug":693,"type":385},"bitwize-music-studio","claude-ai-music-skills",{"evaluate":695,"extract":704},{"promptVersionExtension":26,"promptVersionScoring":27,"score":62,"tags":696,"targetMarket":35,"tier":36},[697,355,698,699,700,701,702,592,703,602],"music-generation","suno","audio-production","workflow","lyrics","mastering","python",{"commitSha":38,"license":705,"plugin":706},"CC0-1.0",{"mcpCount":11,"provider":408,"skillCount":707},54,{"parentExtensionId":709,"repoId":710},"k17bfryzkzywswf1bkgrtch16d86n8t9","kd70cgrajsrnk5gmq60rhq30zd86nyc0",[355,699,602,592,701,702,697,703,698,700],{"evaluatedAt":713,"extractAt":714,"updatedAt":713},1778683131031,1778683100520,{"_creationTime":716,"_id":717,"community":718,"display":719,"identity":724,"providers":728,"relations":734,"tags":737,"workflow":738},1778699018122.7732,"k17a0wf6mk0f48w5xah6yx5dts86n1xj",{"reviewCount":11},{"description":720,"installMethods":721,"name":722,"sourceUrl":723},"Performance analysis, test coverage review, and AI-powered code quality assessment",{"claudeCode":722},"performance-testing-review","https://github.com/wshobson/agents",{"basePath":725,"githubOwner":726,"githubRepo":727,"locale":395,"slug":722,"type":385},"plugins/performance-testing-review","wshobson","agents",{"evaluate":729,"extract":733},{"promptVersionExtension":26,"promptVersionScoring":27,"score":87,"tags":730,"targetMarket":35,"tier":36},[731,732,666,137,355],"code-quality","performance-analysis",{"commitSha":38,"license":39},{"parentExtensionId":735,"repoId":736},"k17cywe30jfsfw3cdpncjfn8y186nvyw","kd74de64zj0axtg5b8t7eqqe2x86nske",[355,731,137,732,666],{"evaluatedAt":739,"extractAt":740,"updatedAt":739},1778699546351,1778699018122,{"_creationTime":742,"_id":743,"community":744,"display":745,"identity":749,"providers":751,"relations":760,"tags":761,"workflow":762},1778690773482.4858,"k175rwqsqyx8atwtz5cs5b3fpx86m84e",{"reviewCount":11},{"description":746,"installMethods":747,"name":748,"sourceUrl":630},"Train or fine-tune sentence-transformers models across all three architectures: SentenceTransformer (bi-encoder embeddings), CrossEncoder (rerankers), and SparseEncoder (SPLADE). Covers loss selection, hard-negative mining, evaluators, distillation, LoRA, Matryoshka, and Hugging Face Hub publishing.",{"claudeCode":748},"train-sentence-transformers",{"basePath":750,"githubOwner":633,"githubRepo":634,"locale":395,"slug":748,"type":385},"skills/train-sentence-transformers",{"evaluate":752,"extract":759},{"promptVersionExtension":26,"promptVersionScoring":27,"score":87,"tags":753,"targetMarket":35,"tier":36},[754,360,755,756,757,758],"machine-learning","sentence-transformers","model-training","embeddings","reranking",{"commitSha":38},{"parentExtensionId":645,"repoId":646},[757,754,756,360,758,755],{"evaluatedAt":763,"extractAt":650,"updatedAt":763},1778691173389,{"_creationTime":765,"_id":766,"community":767,"display":768,"identity":774,"providers":776,"relations":784,"tags":788,"workflow":789},1778692081464.8755,"k171af3ep240f5k9cjztfs0tvd86mg4t",{"reviewCount":11},{"description":769,"installMethods":770,"name":772,"sourceUrl":773},"超压缩通信模式。通过像穴居人一样说话，将 token 数量减少约 75%，同时保持完整技术准确性。",{"claudeCode":771},"caveman","Caveman","https://github.com/juliusbrussee/caveman",{"basePath":384,"githubOwner":775,"githubRepo":771,"locale":22,"slug":771,"type":385},"juliusbrussee",{"evaluate":777,"extract":781},{"promptVersionExtension":26,"promptVersionScoring":27,"score":87,"tags":778,"targetMarket":35,"tier":36},[779,606,356,138,780],"compression","productivity",{"commitSha":38,"license":39,"plugin":782},{"mcpCount":11,"provider":408,"skillCount":783},6,{"parentExtensionId":785,"repoId":786,"translatedFrom":787},"k17dcfwpwpzjc43zt2qawvft3h86mmse","kd7890f0sfy0ymf1pakpx7vqch86nmew","k170g496f5ze5hv8xa2e7b91ex86mz9m",[779,138,606,780,356],{"evaluatedAt":790,"extractAt":791,"updatedAt":792},1778691819811,1778691765524,1778692081464]