[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"similar-foryourhealth111-pixel--Vibe-Skills":3,"tool-foryourhealth111-pixel--Vibe-Skills":64},[4,17,27,35,43,56],{"id":5,"name":6,"github_repo":7,"description_zh":8,"stars":9,"difficulty_score":10,"last_commit_at":11,"category_tags":12,"status":16},3808,"stable-diffusion-webui","AUTOMATIC1111\u002Fstable-diffusion-webui","stable-diffusion-webui 是一个基于 Gradio 构建的网页版操作界面，旨在让用户能够轻松地在本地运行和使用强大的 Stable Diffusion 图像生成模型。它解决了原始模型依赖命令行、操作门槛高且功能分散的痛点，将复杂的 AI 绘图流程整合进一个直观易用的图形化平台。\n\n无论是希望快速上手的普通创作者、需要精细控制画面细节的设计师，还是想要深入探索模型潜力的开发者与研究人员，都能从中获益。其核心亮点在于极高的功能丰富度：不仅支持文生图、图生图、局部重绘（Inpainting）和外绘（Outpainting）等基础模式，还独创了注意力机制调整、提示词矩阵、负向提示词以及“高清修复”等高级功能。此外，它内置了 GFPGAN 和 CodeFormer 等人脸修复工具，支持多种神经网络放大算法，并允许用户通过插件系统无限扩展能力。即使是显存有限的设备，stable-diffusion-webui 也提供了相应的优化选项，让高质量的 AI 艺术创作变得触手可及。",162132,3,"2026-04-05T11:01:52",[13,14,15],"开发框架","图像","Agent","ready",{"id":18,"name":19,"github_repo":20,"description_zh":21,"stars":22,"difficulty_score":23,"last_commit_at":24,"category_tags":25,"status":16},1381,"everything-claude-code","affaan-m\u002Feverything-claude-code","everything-claude-code 是一套专为 AI 编程助手（如 Claude Code、Codex、Cursor 等）打造的高性能优化系统。它不仅仅是一组配置文件，而是一个经过长期实战打磨的完整框架，旨在解决 AI 代理在实际开发中面临的效率低下、记忆丢失、安全隐患及缺乏持续学习能力等核心痛点。\n\n通过引入技能模块化、直觉增强、记忆持久化机制以及内置的安全扫描功能，everything-claude-code 能显著提升 AI 在复杂任务中的表现，帮助开发者构建更稳定、更智能的生产级 AI 代理。其独特的“研究优先”开发理念和针对 Token 消耗的优化策略，使得模型响应更快、成本更低，同时有效防御潜在的攻击向量。\n\n这套工具特别适合软件开发者、AI 研究人员以及希望深度定制 AI 工作流的技术团队使用。无论您是在构建大型代码库，还是需要 AI 协助进行安全审计与自动化测试，everything-claude-code 都能提供强大的底层支持。作为一个曾荣获 Anthropic 黑客大奖的开源项目，它融合了多语言支持与丰富的实战钩子（hooks），让 AI 真正成长为懂上",140436,2,"2026-04-05T23:32:43",[13,15,26],"语言模型",{"id":28,"name":29,"github_repo":30,"description_zh":31,"stars":32,"difficulty_score":23,"last_commit_at":33,"category_tags":34,"status":16},2271,"ComfyUI","Comfy-Org\u002FComfyUI","ComfyUI 是一款功能强大且高度模块化的视觉 AI 引擎，专为设计和执行复杂的 Stable Diffusion 图像生成流程而打造。它摒弃了传统的代码编写模式，采用直观的节点式流程图界面，让用户通过连接不同的功能模块即可构建个性化的生成管线。\n\n这一设计巧妙解决了高级 AI 绘图工作流配置复杂、灵活性不足的痛点。用户无需具备编程背景，也能自由组合模型、调整参数并实时预览效果，轻松实现从基础文生图到多步骤高清修复等各类复杂任务。ComfyUI 拥有极佳的兼容性，不仅支持 Windows、macOS 和 Linux 全平台，还广泛适配 NVIDIA、AMD、Intel 及苹果 Silicon 等多种硬件架构，并率先支持 SDXL、Flux、SD3 等前沿模型。\n\n无论是希望深入探索算法潜力的研究人员和开发者，还是追求极致创作自由度的设计师与资深 AI 绘画爱好者，ComfyUI 都能提供强大的支持。其独特的模块化架构允许社区不断扩展新功能，使其成为当前最灵活、生态最丰富的开源扩散模型工具之一，帮助用户将创意高效转化为现实。",107662,"2026-04-03T11:11:01",[13,14,15],{"id":36,"name":37,"github_repo":38,"description_zh":39,"stars":40,"difficulty_score":23,"last_commit_at":41,"category_tags":42,"status":16},3704,"NextChat","ChatGPTNextWeb\u002FNextChat","NextChat 是一款轻量且极速的 AI 助手，旨在为用户提供流畅、跨平台的大模型交互体验。它完美解决了用户在多设备间切换时难以保持对话连续性，以及面对众多 AI 模型不知如何统一管理的痛点。无论是日常办公、学习辅助还是创意激发，NextChat 都能让用户随时随地通过网页、iOS、Android、Windows、MacOS 或 Linux 端无缝接入智能服务。\n\n这款工具非常适合普通用户、学生、职场人士以及需要私有化部署的企业团队使用。对于开发者而言，它也提供了便捷的自托管方案，支持一键部署到 Vercel 或 Zeabur 等平台。\n\nNextChat 的核心亮点在于其广泛的模型兼容性，原生支持 Claude、DeepSeek、GPT-4 及 Gemini Pro 等主流大模型，让用户在一个界面即可自由切换不同 AI 能力。此外，它还率先支持 MCP（Model Context Protocol）协议，增强了上下文处理能力。针对企业用户，NextChat 提供专业版解决方案，具备品牌定制、细粒度权限控制、内部知识库整合及安全审计等功能，满足公司对数据隐私和个性化管理的高标准要求。",87618,"2026-04-05T07:20:52",[13,26],{"id":44,"name":45,"github_repo":46,"description_zh":47,"stars":48,"difficulty_score":23,"last_commit_at":49,"category_tags":50,"status":16},2268,"ML-For-Beginners","microsoft\u002FML-For-Beginners","ML-For-Beginners 是由微软推出的一套系统化机器学习入门课程，旨在帮助零基础用户轻松掌握经典机器学习知识。这套课程将学习路径规划为 12 周，包含 26 节精炼课程和 52 道配套测验，内容涵盖从基础概念到实际应用的完整流程，有效解决了初学者面对庞大知识体系时无从下手、缺乏结构化指导的痛点。\n\n无论是希望转型的开发者、需要补充算法背景的研究人员，还是对人工智能充满好奇的普通爱好者，都能从中受益。课程不仅提供了清晰的理论讲解，还强调动手实践，让用户在循序渐进中建立扎实的技能基础。其独特的亮点在于强大的多语言支持，通过自动化机制提供了包括简体中文在内的 50 多种语言版本，极大地降低了全球不同背景用户的学习门槛。此外，项目采用开源协作模式，社区活跃且内容持续更新，确保学习者能获取前沿且准确的技术资讯。如果你正寻找一条清晰、友好且专业的机器学习入门之路，ML-For-Beginners 将是理想的起点。",84991,"2026-04-05T10:45:23",[14,51,52,53,15,54,26,13,55],"数据工具","视频","插件","其他","音频",{"id":57,"name":58,"github_repo":59,"description_zh":60,"stars":61,"difficulty_score":10,"last_commit_at":62,"category_tags":63,"status":16},3128,"ragflow","infiniflow\u002Fragflow","RAGFlow 是一款领先的开源检索增强生成（RAG）引擎，旨在为大语言模型构建更精准、可靠的上下文层。它巧妙地将前沿的 RAG 技术与智能体（Agent）能力相结合，不仅支持从各类文档中高效提取知识，还能让模型基于这些知识进行逻辑推理和任务执行。\n\n在大模型应用中，幻觉问题和知识滞后是常见痛点。RAGFlow 通过深度解析复杂文档结构（如表格、图表及混合排版），显著提升了信息检索的准确度，从而有效减少模型“胡编乱造”的现象，确保回答既有据可依又具备时效性。其内置的智能体机制更进一步，使系统不仅能回答问题，还能自主规划步骤解决复杂问题。\n\n这款工具特别适合开发者、企业技术团队以及 AI 研究人员使用。无论是希望快速搭建私有知识库问答系统，还是致力于探索大模型在垂直领域落地的创新者，都能从中受益。RAGFlow 提供了可视化的工作流编排界面和灵活的 API 接口，既降低了非算法背景用户的上手门槛，也满足了专业开发者对系统深度定制的需求。作为基于 Apache 2.0 协议开源的项目，它正成为连接通用大模型与行业专有知识之间的重要桥梁。",77062,"2026-04-04T04:44:48",[15,14,13,26,54],{"id":65,"github_repo":66,"name":67,"description_en":68,"description_zh":69,"ai_summary_zh":69,"readme_en":70,"readme_zh":71,"quickstart_zh":72,"use_case_zh":73,"hero_image_url":74,"owner_login":75,"owner_name":76,"owner_avatar_url":77,"owner_bio":78,"owner_company":79,"owner_location":79,"owner_email":80,"owner_twitter":79,"owner_website":79,"owner_url":81,"languages":82,"stars":122,"forks":123,"last_commit_at":124,"license":125,"difficulty_score":23,"env_os":126,"env_gpu":126,"env_ram":126,"env_deps":127,"category_tags":137,"github_topics":138,"view_count":23,"oss_zip_url":79,"oss_zip_packed_at":79,"status":16,"created_at":159,"updated_at":160,"faqs":161,"releases":192},2395,"foryourhealth111-pixel\u002FVibe-Skills","Vibe-Skills","Wherever your AI supports skills, VibeSkills works. 340+ governed skills spanning coding, research, automation & creative work.","Vibe-Skills 不仅仅是一个技能集合，它更像是一个专为你的 AI 助手打造的“个人操作系统”。无论你所使用的 AI 平台是否支持扩展技能，Vibe-Skills 都能无缝接入，提供超过 340 种经过严格治理的专业能力模块，涵盖代码编写、学术研究、数据科学、自动化办公及创意创作等多个领域。\n\n在日常使用中，用户往往面临 AI 回答泛泛而谈或执行流程混乱的痛点。Vibe-Skills 通过其独特的“规范运行时”机制解决了这一问题：当你发出指令时，系统不会盲目执行，而是自动经历“需求澄清→任务规划→智能路由→执行验证→最终交付”的严谨流程。其内置的“规范路由器”能根据任务类型，自动调用最合适的专家技能（如代码审查、测试驱动开发指南等），确保多技能协作时无冲突且高效。\n\n这款工具非常适合开发者、研究人员、设计师以及希望深度挖掘 AI 潜力的进阶用户。对于开发者而言，它是强大的工程辅助伙伴；对于研究人员和创意工作者，它则提供了结构化的工作流支持。只需简单安装并输入 `\u002Fvibe` 指令，即可激活这套智能系统，让 AI 从单纯的聊天机器人进化为真正具备记忆力和执行力的全能工作搭档。","\u003Cdiv align=\"right\">\n  \u003Cb>🇬🇧 English\u003C\u002Fb>  |  \u003Ca href=\".\u002FREADME.zh.md\">🇨🇳 中文\u003C\u002Fa>\n\u003C\u002Fdiv>\n\n\u003Cbr\u002F>\n\n\u003Cdiv align=\"center\">\n\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\">\n  \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fforyourhealth111-pixel_Vibe-Skills_readme_f9e121099867.png\" alt=\"VibeSkills Typing Logo\" \u002F>\n\u003C\u002Fa>\n\n\u003Cbr\u002F>\n\n\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fforyourhealth111-pixel_Vibe-Skills_readme_947a6dc2a01f.png\" width=\"260px\" alt=\"VibeSkills Logo\"\u002F>\n\n\u003Cbr\u002F>\u003Cbr\u002F>\n\n### More than a skill collection — your **personal AI operating system**\n\nIf your AI supports skills, VibeSkills works. 340+ skills spanning coding, research, data science & creative work.\n\n\u003Cbr\u002F>\n\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\u002Fstargazers\">\n  \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fforyourhealth111-pixel\u002FVibe-Skills?style=for-the-badge&logo=github&color=7B61FF&label=STARS\" alt=\"stars\">\n\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\u002Fnetwork\u002Fmembers\">\n  \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fforks\u002Fforyourhealth111-pixel\u002FVibe-Skills?style=for-the-badge&logo=git&color=45a1ff&label=FORKS\" alt=\"forks\">\n\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\u002Fpulse\">\n  \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Flast-commit\u002Fforyourhealth111-pixel\u002FVibe-Skills?style=for-the-badge&logo=git-lfs&color=32CD32&label=MOMENTUM\" alt=\"last commit\">\n\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgitcgr.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\">\n  \u003Cimg src=\"https:\u002F\u002Fgitcgr.com\u002Fbadge\u002Fforyourhealth111-pixel\u002FVibe-Skills.svg\" alt=\"gitcgr\" \u002F>\n\u003C\u002Fa>\n\n\u003Cbr\u002F>\u003Cbr\u002F>\n\n\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fforyourhealth111-pixel_Vibe-Skills_readme_510b79be381e.png\" alt=\"Visitors\">\n \n\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FArchitecture-VCO_Runtime-orange?style=for-the-badge\" alt=\"Arch\">\n \n\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FSkills-340%2B-blueviolet?style=for-the-badge\" alt=\"Skills Count\">\n\n\u003Cbr\u002F>\u003Cbr\u002F>\n\n🧠 Planning · 🛠️ Engineering · 🤖 AI · 🔬 Research · 🎨 Creation\n\n\u003Cbr\u002F>\n\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\u002Fblob\u002Fmain\u002Fdocs\u002Finstall\u002Fone-click-install-release-copy.en.md\">\n  \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002F⚡_Get_Started-7B61FF?style=for-the-badge\" alt=\"Install\">\n\u003C\u002Fa>\n \n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\u002Fblob\u002Fmain\u002Fdocs\u002Fquick-start.en.md\">\n  \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002F📖_Quick_Start-2d3748?style=for-the-badge\" alt=\"Docs\">\n\u003C\u002Fa>\n \n\u003Ca href=\".\u002FREADME.zh.md\">\n  \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002F🇨🇳_中文-45a1ff?style=for-the-badge\" alt=\"Chinese\">\n\u003C\u002Fa>\n\n\u003Cbr\u002F>\u003Cbr\u002F>\n\n\u003Ckbd>Install\u003C\u002Fkbd>  → \n\u003Ckbd>\u002Fvibe or $vibe\u003C\u002Fkbd>  → \n\u003Ckbd>Smart Routing\u003C\u002Fkbd>  → \n\u003Ckbd>M \u002F L \u002F XL Execution\u003C\u002Fkbd>  → \n\u003Ckbd>Governance Verification\u003C\u002Fkbd>  → \n\u003Ckbd>✅ Delivery\u003C\u002Fkbd>\n\n\u003C\u002Fdiv>\n\n## 📋 Table of Contents\n\n- [What makes it different](#-what-makes-it-different)\n- [Who is it for](#-who-is-it-for)\n- [Intelligent Routing](#-intelligent-routing-how-340-skills-collaborate-without-conflict)\n- [Memory System](#-memory-system-ai-that-truly-remembers)\n- [Full Capability Map](#-full-capability-map-your-all-in-one-workbench)\n- [Installation & Management](#️-installation--skills-management)\n- [Getting Started](#-getting-started)\n\n\n\u003Cdetails>\n\u003Csummary>\u003Cb>🔑 New here? Quick glossary of key terms (click to expand)\u003C\u002Fb>\u003C\u002Fsummary>\n\n\u003Cbr\u002F>\n\n| Term | Plain-English Meaning |\n|:---|:---|\n| **VibeSkills \u002F VCO** | This project. VCO = Vibe Code Orchestrator — the runtime engine behind the skills. |\n| **Skill** | A focused capability module (e.g., `tdd-guide`, `code-review`). Think of skills as expert assistants the system calls on demand. |\n| **Governed runtime** | When you invoke `\u002Fvibe`, the system follows a structured process — clarify → plan → execute → verify — instead of diving in blindly. |\n| **Canonical Router** | The internal logic that decides which skill to activate for your task. Just invoke `\u002Fvibe` and let it route automatically. |\n| **M \u002F L \u002F XL grade** | Task complexity level. M = quick focused task, L = multi-step task, XL = large task with parallel work. Automatically selected. |\n| **Frozen requirement** | Once you confirm the plan, it is \"frozen\" — the system will not silently change scope mid-task. |\n| **Root \u002F Child lane** | In XL tasks, there is a \"root\" coordinator and \"child\" worker agents. Prevents conflicting outputs from parallel agents. |\n| **Proof bundle** | Evidence that a task was actually completed correctly — test results, output, verification logs. |\n\n\u003C\u002Fdetails>\n\n> [!IMPORTANT]\n> ### 🎯 Core Vision\n>\n> VibeSkills evolves with the times — ensuring it stays genuinely useful while **dramatically lowering the barrier to cutting-edge vibecoding technology**, eliminating the cognitive anxiety and steep learning curve that comes with new AI tools.\n>\n> **Whether or not you have a programming background, you can directly harness the most advanced AI capabilities with minimal effort.**\n> Productivity gains from AI should be available to everyone.\n\n\u003Cbr\u002F>\n\n---\n\n\n## ✨ What makes it different?\n\n> Traditional skill repos answer: _\"What tools do I have?\"_\n> **VibeSkills tackles the core pain point of heavy AI users: _\"How do I manage and invoke large numbers of Skills, and get work done efficiently and reliably?\"_**\n\n\u003Csub>Works with **Claude Code** · **Codex** · **Windsurf** · **OpenClaw** · **OpenCode** · **Cursor** and any AI environment that supports the Skills protocol. Native **MCP** compatibility.\u003C\u002Fsub>\n\n\u003Cbr\u002F>\n\n\u003Cdiv align=\"center\">\n\n| ❌  Traditional Pain Points (you've probably felt these) | ✅  VibeSkills Solutions (what we've built) |\n|:---|:---|\n| **Skills never activate**: Hundreds of capabilities in the repo, but AI rarely remembers to use them — activation rate is extremely low. | **🧠 Intelligent Routing**: The system automatically routes to the right skill based on context — no need to memorize a skill list. |\n| **Blind execution**: AI dives in without clarifying requirements — fast but off-target, projects gradually become black boxes. | **🧭 Governed Workflow**: Clarify → Verify → Trace is enforced in a unified process; every step is auditable. |\n| **Conflicting tools**: Lack of coordination between plugins and workflows leads to environment pollution or infinite loops. | **🧩 Global Governance**: 129 contract rules define safety boundaries and fallback mechanisms for long-term stability. |\n| **Messy workspace**: After extended use, repos become cluttered; new Agents miss project details when taking over, causing handoff gaps. | **📁 Semantic Directory Governance**: Fixed-architecture file storage so any new AI conversation instantly understands the project context. |\n| **AI bad habits**: Deletes main files while clearing backups; writes silent fallbacks then confidently claims \"it's done\". | **🛡️ Built-in Safety Rules**: Governed execution blocks dangerous bulk deletion and blind recursive wipes by default; fallback mechanisms must always show explicit warnings. |\n| **Manual workflow discipline**: Users must maintain their own AI collaboration process from experience — high learning cost. | **🚦 Framework-guided end-to-end**: Requirements → Plan → Multi-agent execution → Automated test iteration — fully managed. |\n| **Skill dispatch chaos in multi-agent runs**: Hard to assign the right skills to each agent for different tasks. | **🤖 Automatic Skill Dispatch**: Multi-agent workflows automatically assign the corresponding Skills to each Agent's task. |\n\n\u003C\u002Fdiv>\n\n\u003Cbr\u002F>\n\n---\n\n\n## 👥 Who is it for?\n\n_Which of those pain points hit home? Find your position — what comes next will land harder._\n\n\u003Cdetails>\n\u003Csummary>Is this for you? Click to expand\u003C\u002Fsummary>\n\n\u003Cbr\u002F>\n\n\u003Cdiv align=\"center\">\n\n| Audience | Description |\n|:---:|:---|\n| 🎯 **Users who need reliable delivery** | Want AI to be a dependable partner, not a runaway horse |\n| ⚡ **Power users heavily relying on AI\u002FAgents** | Need a unified foundation to support large-scale workflows |\n| 🏢 **Small teams with high standardization needs** | Want AI workflows to be more maintainable and transferable |\n| 😩 **Practitioners exhausted by skill sprawl** | Already tired of tool hunting — just want a ready-to-use solution |\n\n\u003C\u002Fdiv>\n\n> _If you're looking for a single small script, this may be overkill. But if you want to use AI more reliably, smoothly, and sustainably — this is your indispensable foundation._\n\n\u003C\u002Fdetails>\n\n\u003Cbr\u002F>\n\n---\n\n\n## 🔀 Intelligent Routing: How 340+ Skills Collaborate Without Conflict\n\n_You know this is for you. Next question: 340+ skills in one system — how do they stay out of each other's way?_\n\nWith 340+ skills, you might wonder: _\"Won't similar skills conflict? How does the system know which one to use?\"_\n\n### How routing works\n\nVibeSkills uses a **Canonical Router** as the single authoritative routing decision center:\n\n```mermaid\ngraph LR\n    A[User Task] --> B{Canonical Router}\n    B --> C[Intent Recognition]\n    C --> D[Keyword Extraction]\n    D --> E[Skill Matching]\n    E --> F[Conflict Detection]\n    F --> G[Priority Ranking]\n    G --> H[Routing Decision]\n    H --> I[Execute Skill]\n\n    style B fill:#7B61FF,stroke:#fff,stroke-width:2px,color:#fff\n    style F fill:#FF9800,stroke:#fff,stroke-width:2px,color:#fff\n```\n\nVibeSkills follows a `Clarify ➔ Plan ➔ Execute ➔ Verify` governed workflow to ensure every task goes through complete quality control:\n\n- **Requirements Clarification**: Skills like `speckit-clarify` define clear boundaries and acceptance criteria\n- **Architecture Planning**: Skills like `aios-architect` design the implementation path\n- **Execution Layer**: 340+ skills called on demand to complete the actual work\n- **Quality Verification**: Skills like `tdd-guide` and `code-review` ensure delivery quality\n\n---\n\n### Why this design?\n\nTraditional skill repos let AI \"freely choose\" — the result:\n\n- ❌ AI can't remember what skills exist\n- ❌ Similar skills conflict with each other\n- ❌ Execution paths are unpredictable\n\nVibeSkills routing guarantees:\n\n- ✅ **Determinism**: Same task always follows the same routing logic\n- ✅ **Traceability**: Every routing decision has a clear rationale\n- ✅ **Control**: Users can override default routing via explicit invocation (e.g. `\u002Fvibe`)\n- ✅ **Stability**: 129 governance rules prevent conflicts and divergence\n\n---\n\n### M \u002F L \u002F XL Execution Levels\n\nAfter selecting the primary skill, the router also automatically determines the execution level based on task complexity:\n\n\u003Cdiv align=\"center\">\n\n| Level | Use Case | Characteristics |\n|:---:|:---|:---|\n| **M** | Narrow-scope work with clear boundaries | Single-agent, token-efficient, fast response |\n| **L** | Medium complexity requiring design, planning, and review | Native serial execution by planned steps; bounded delegated units only when explicitly planned |\n| **XL** | Large tasks — parallelizable, long-running, multi-agent wave execution | Wave-sequential orchestration with step-level bounded parallelism for independent units only |\n\n\u003C\u002Fdiv>\n\n> The system automatically selects the level after requirements clarification, before plan execution. Users only need to invoke `\u002Fvibe` or `$vibe`.\n>\n> When the system calls a specialist skill internally (like `tdd-guide` or `code-review`), it is always scoped to a specific phase — they assist without taking over the overall coordination. In XL tasks with multiple agents, worker agents (child lanes) can suggest specialist help, but the coordinator (root) approves it before execution.\n>\n> You can also express an explicit preference:\n> ```text\n> Please execute this task according to the plan, launching XL-level workflow \u002Fvibe\n> ```\n\n---\n\n\u003Cdetails>\n\u003Csummary>\u003Cb>🔍 Routing FAQ (click to expand)\u003C\u002Fb>\u003C\u002Fsummary>\n\n\u003Cbr\u002F>\n\n**One route or multiple per task?**\n\nCore principle: A task typically routes to one primary skill, but that skill can invoke others as sub-processes.\n\n- **Single primary route**: The Canonical Router selects **the single best-matching primary skill**\n- **Skill composition**: The primary skill can invoke others as needed during execution (e.g. `vibe` can invoke `speckit-clarify`, `aios-architect`, etc.)\n- **Governed coordination**: Multi-skill collaboration is controlled by governance rules, not arbitrary combinations\n\n\u003Cbr\u002F>\n\n**How are conflicts between similar skills handled?**\n\nWhen multiple skills appear capable of completing a task, the router avoids conflicts through:\n\n1. **Priority rules**: Each skill has a clear priority and applicable scenario\n2. **Context matching**: Analyzes task complexity, multi-phase needs, and explicit user preferences\n3. **Mutual exclusion rules**: 129 rules include exclusion rules preventing conflicting combinations\n4. **Graceful degradation**: When the preferred skill is unavailable, fallback by priority — no infinite loops\n\n\u003Cbr\u002F>\n\n**Will too many options cause token explosion?**\n\nNo. Routing doesn't dump all options into the model — it uses a smart trigger mechanism:\n\n```\nUser command → AI-assisted governance extracts intent keywords → keywords trigger skill routing\n```\n\nThe governance framework adds ~30k initial context overhead, but does not cause token explosion.\n\n\u003Cbr\u002F>\n\n**Real example: User says \"Help me refactor this project\"**\n\n1. Intent recognition → Complex refactoring task\n2. Keyword extraction → refactor, project, code quality\n3. Skill matching → `vibe` \u002F `autonomous-builder` \u002F `systematic-debugging`\n4. Routing decision → Choose `vibe` (refactoring needs multi-phase: clarify → plan → execute → verify)\n\n\u003C\u002Fdetails>\n\n\u003Cbr\u002F>\n\n---\n\n\n## 🧠 Memory System: AI That Truly Remembers\n\n_Routing solves \"which skill\". But there's a deeper question: when the conversation ends, does AI remember you?_\n\nSound familiar?\n\n\u003Cdiv align=\"center\">\n\n| ❌ Pain Point | ✅ VibeSkills Solution | Component |\n|:---|:---|:---:|\n| Re-explaining project context every new session | Architecture decisions & conventions auto-loaded on startup | `Serena` |\n| AI hits the same bugs again; insights vanish with context | One sentence saves to Obsidian + GitHub permanently | `knowledge-steward` |\n| Long tasks — AI gradually \"forgets\" early context | In-session semantic vector cache, instant retrieval | `ruflo` |\n| Cross-project knowledge can't accumulate | Entity relationship graphs grow richer over time | `Cognee` |\n| Long task interrupted — hard to hand off to new agent | Auto-folds into working + tool + evidence memory | `deepagent-memory-fold` |\n\n\u003C\u002Fdiv>\n\n\u003Cbr\u002F>\n\n\u003Cdetails>\n\u003Csummary>\u003Cb>📐 Expand: Four-Tier Architecture, Memory Skills & Governance Rules\u003C\u002Fb>\u003C\u002Fsummary>\n\n\u003Cbr\u002F>\n\nVibeSkills builds a **four-tier memory system** — one authoritative component per memory need:\n\n| Tier | Component | Scope | Core Purpose |\n|:---:|:---:|:---:|:---|\n| **L1 Session** | `state_store` | Current session | Execution progress, intermediate results, temp state — always-on \"workbench\" |\n| **L2 Project** | `Serena` | Current project | Architecture decisions, conventions — written only after explicit user confirmation |\n| **L3 Short-term Semantic** | `ruflo` | Intra-session | Vector cache for fast context retrieval within long-running tasks |\n| **L4 Long-term Graph** | `Cognee` | Cross-session | Entity linking, relationship graphs, long-horizon knowledge accumulation |\n\n> **Optional extensions**: `mem0` as a personal preference backend (opt-in); `Letta` provides memory block mapping vocabulary — neither replaces the four canonical tiers.\n\n\u003Cbr\u002F>\n\n**Three Dedicated Memory Skills**\n\n| Skill | Role | Trigger |\n|:---:|:---|:---|\n| `knowledge-steward` | **Knowledge Keeper**: Saves insights, bug fixes, and prompts to Obsidian + GitHub permanently | \"save this prompt\" \u002F \"log this bug\" \u002F \"save this insight\" |\n| `digital-brain` | **Second Brain**: Structured personal knowledge base — identity, content, network, retrospectives | Invoke directly; ideal for a personal knowledge OS |\n| `deepagent-memory-fold` | **Context Fold**: Compresses large context into structured working\u002Ftool\u002Fevidence memory for seamless handoff | Triggers at context limit or manually |\n\n\u003Cbr\u002F>\n\n**Governance**: Single source of truth (no dual-track) · Explicit write only (`Serena` requires confirmation) · `episodic-memory` permanently disabled · `mem0` limited to personal preferences · Kill switch on every external backend\n\n\u003C\u002Fdetails>\n\n\n---\n\n\n## ✦ Full Capability Map: Your All-in-One Workbench\n\n_Routing + memory form the dispatch nervous system. Here's the full capability map they power — end to end._\n\nUnrolled across a \"real workflow\", VibeSkills has laid out a complete **end-to-end capability chain**:\n\n\u003Cbr\u002F>\n\n\u003Cdiv align=\"center\">\n\n| Domain | Coverage | Representative Engines |\n|:---|:---|:---|\n| **💡 Requirements & Clarification** | No more black-box starts: turn vague ideas into clearly-bounded, verifiable problem definitions | `brainstorming`, `speckit-clarify` |\n| **📋 Planning & Breakdown** | Decompose ambitious goals into specs, plans, tasks, milestones, and execution flows | `writing-plans`, `speckit-specify`, `aios-po` |\n| **🏗️ Architecture & Tech Selection** | Design frontend\u002Fbackend boundaries, APIs, data layers, deployment, and tech stack comparisons | `aios-architect`, `architecture-patterns` |\n| **💻 Development & Implementation** | New features, scaffolding, engineering integration, and precise cross-file implementation | `autonomous-builder`, `speckit-implement` |\n| **🔧 Debugging & Refactoring** | Beyond surface patches: locate errors, analyze root causes, restore project-level maintainability | `error-resolver`, `systematic-debugging` |\n| **🛡️ Testing & Quality Control** | Unit tests, regression verification, quality gates — mandatory verification before completion | `tdd-guide`, `aios-qa`, `code-review` |\n| **🚀 Collaboration & Release** | Handle Issues\u002FPRs, fix CI, process reviews, and automated deployment | `aios-devops`, `gh-fix-ci`, `vercel-deploy` |\n| **🤖 Compound Workflows** | Freeze requirements, dispatch tasks, multi-Agent coordination, execution tracing, env cleanup | `vibe`, `swarm_*`, `hive-mind-advanced` |\n| **🔌 External Ecosystem** | Bridge browsers, web scraping, design files, third-party services, and context memory | `mcp-integration`, `playwright`, `scrapling` |\n| **📊 Data & AI Engineering** | EDA, cleaning & stats, to model training, RAG retrieval, and experiment tracking | `senior-ml-engineer`, `statistical-analysis` |\n| **🔬 Research & Life Sciences** | **Core strength**: literature review, bioinformatics, single-cell analysis, drug discovery | `literature-review`, `biopython`, `scanpy` |\n| **📐 Math & Scientific Computing** | Symbolic derivation, Bayesian modeling, multi-objective optimization, simulation, quantum computing | `sympy`, `pymc-bayesian-modeling`, `qiskit` |\n| **🎨 Multimedia & Presentation** | Interactive charts, publication-quality figures, slides, audio\u002Fvideo production | `plotly`, `generate-image`, `video-studio` |\n\n\u003C\u002Fdiv>\n\n\u003Cbr\u002F>\n\n\u003Cdetails>\n\u003Csummary>\u003Cb>👉 Expand: Explore the complete 340+ full-stack capability matrix\u003C\u002Fb>\u003C\u002Fsummary>\n\n\u003Cbr\u002F>\n\n> 💡 **Why governance matters**: The vast skill library below is not a collection of isolated scripts — it is an ecosystem governed by the VCO runtime. Through domain matrix classification, the system automatically invokes the right tool at the right context node, without requiring you to manually search through skills.\n\n---\n\n### 🧠 Requirements, Planning & Product Management\n\n> **Turn big ideas into actionable plans**: requirement insights, problem definition, Sprint planning, task breakdown, and constraint collection. Ensure direction is clear, boundaries are defined, and milestones are verifiable before writing a single line of code.\n\n`.system`, `aios-pm`, `aios-po`, `aios-sm`, `aios-squad-creator`, `aios-ux-design-expert`, `brainstorming`, `create-plan`, `designing-experiments`, `planning-with-files`, `shared-templates`, `speckit-analyze`, `speckit-checklist`, `speckit-clarify`, `speckit-constitution`, `speckit-plan`, `speckit-specify`, `speckit-tasks`, `speckit-taskstoissues`, `subagent-driven-development`, `think-harder`, `treatment-plans`, `ux-researcher-designer`, `writing-plans`\n\n---\n\n### 🛠️ Software Engineering & Architecture\n\n> **The true engineering foundation**: from scaffolding, cross-file changes, API design to microservice architecture evaluation. Not just code output — also context memory, toolchain orchestration, and multi-phase intelligent Agent coordination.\n\n`aios-architect`, `aios-dev`, `aios-master`, `architecture-patterns`, `autonomous-builder`, `cancel-ralph`, `coding-tutor`, `context-fundamentals`, `context-hunter`, `cs-foundations`, `deepagent-memory-fold`, `deepagent-toolchain-plan`, `evaluating-code-models`, `get-available-resources`, `hive-mind-advanced`, `local-vco-roles`, `nowait-reasoning-optimizer`, `prompt-lookup`, `ralph-loop`, `skill-creator`, `skill-lookup`, `spec-kit-vibe-compat`, `speckit-implement`, `superclaude-framework-compat`, `theme-factory`, `vibe`, `webthinker-deep-research`\n\n---\n\n### 🔧 Debugging, Testing & Quality Assurance\n\n> **Guarding the lifeline of code and systems**: unit tests, root cause analysis, dependency conflict resolution, security vulnerability reviews, and a complete TDD guide — ensuring systems never enter a \"breaks after every change\" black-box state.\n\n`aios-qa`, `build-error-resolver`, `code-review`, `code-review-excellence`, `code-reviewer`, `data-quality-checker`, `data-quality-frameworks`, `debugging-strategies`, `deslop`, `detecting-performance-regressions`, `error-resolver`, `evals-context`, `experiment-failure-analysis`, `generating-test-reports`, `ml-data-leakage-guard`, `performance-testing`, `property-based-testing`, `providing-performance-optimization-advice`, `receiving-code-review`, `requesting-code-review`, `reviewing-code`, `security-best-practices`, `security-ownership-map`, `security-reviewer`, `security-threat-model`, `systematic-debugging`, `tdd-guide`, `verification-before-completion`, `verification-quality-assurance`, `windows-hook-debugging`\n\n---\n\n### 📊 Data Analysis & Statistical Modeling\n\n> **Let data tell the truth**: a one-stop data processing engine from data cleaning, missing value handling, and exploratory analysis (EDA) to advanced statistical testing, regression models, and time series forecasting.\n\n`aios-data-engineer`, `anomaly-detector`, `correlation-analyzer`, `dask`, `data-artist`, `data-exploration-visualization`, `data-normalization-tool`, `detecting-data-anomalies`, `excel-analysis`, `exploratory-data-analysis`, `feature-importance-analyzer`, `geopandas`, `hypothesis-testing`, `metric-calculator`, `networkx`, `performing-causal-analysis`, `performing-regression-analysis`, `polars`, `preprocessing-data-with-automated-pipelines`, `regression-analysis-helper`, `running-clustering-algorithms`, `scientific-data-preprocessing`, `splitting-datasets`, `spreadsheet`, `statistical-analysis`, `statistics-math`, `statsmodels`, `usfiscaldata`, `vaex`, `xlsx`\n\n---\n\n### 🤖 Machine Learning & AI Engineering\n\n> **Full-stack AI model development**: beyond just calling APIs — feature engineering, model training, fine-tuning, interpretability (SHAP), large model evaluation (Evals), and reinforcement learning workflows.\n\n`LQF_Machine_Learning_Expert_Guide`, `aeon`, `datamol`, `deepchem`, `embedding-strategies`, `engineering-features-for-machine-learning`, `evaluating-llms-harness`, `evaluating-machine-learning-models`, `explaining-machine-learning-models`, `geniml`, `ml-pipeline-workflow`, `openai-knowledge`, `pufferlib`, `pytorch-lightning`, `scikit-learn`, `scikit-survival`, `senior-computer-vision`, `senior-data-scientist`, `senior-ml-engineer`, `senior-prompt-engineer`, `shap`, `similarity-search-patterns`, `sparse-autoencoder-training`, `stable-baselines3`, `tensorboard`, `timesfm-forecasting`, `torch-geometric`, `torch_geometric`, `torchdrug`, `training-machine-learning-models`, `transformer-lens-interpretability`, `transformers`, `umap-learn`, `unsloth`, `weights-and-biases`\n\n---\n\n### 🧬 Life Sciences & Bioinformatics\n\n> **A formidable interdisciplinary powerhouse**: single-cell sequencing analysis, protein structure folding, drug molecule discovery, genomics alignment — seamlessly integrated with cloud-based biology lab systems.\n\n`adaptyv`, `alphafold-database`, `anndata`, `arboreto`, `benchling-integration`, `biopython`, `bioservices`, `cellxgene-census`, `cobrapy`, `deeptools`, `diffdock`, `dnanexus-integration`, `esm`, `etetoolkit`, `flowio`, `gene-database`, `gget`, `ginkgo-cloud-lab`, `gtars`, `histolab`, `imaging-data-commons`, `labarchive-integration`, `lamindb`, `latchbio-integration`, `matchms`, `medchem`, `molfeat`, `neurokit2`, `neuropixels-analysis`, `omero-integration`, `opentrons-integration`, `pathml`, `protocolsio-integration`, `pydeseq2`, `pydicom`, `pyhealth`, `pylabrobot`, `pyopenms`, `pysam`, `pytdc`, `rdkit`, `scanpy`, `scikit-bio`, `scvi-tools`, `tiledbvcf`\n\n---\n\n### 🔬 Scientific Computing & Mathematical Logic\n\n> **Precise derivation and complex system simulation**: symbolic math, Bayesian probabilistic programming, quantum computing simulation, multi-objective optimization, and rigorous propositional logic and mathematical proof assistance.\n\n`astropy`, `cirq`, `dialectic`, `fluidsim`, `gradient-methods`, `math`, `math-model-selector`, `math-tools`, `mathematical-logic-expert`, `matlab`, `pennylane`, `pymatgen`, `pymc`, `pymc-bayesian-modeling`, `pymoo`, `propositional-logic`, `qiskit`, `qutip`, `rowan`, `simpy`, `sympy`, `xan`\n\n---\n\n### 📚 Scientific Literature & Academic Writing\n\n> **The expressway for academic productivity**: precise search across dozens of databases (PubMed, arXiv, etc.), systematic review matrix organization, citation management, and the complete publication pipeline from drafting to peer review.\n\n`bgpt-paper-search`, `biorxiv-database`, `brenda-database`, `chembl-database`, `citation-management`, `clinical-decision-support`, `clinical-reports`, `clinicaltrials-database`, `clinpgx-database`, `clinvar-database`, `comprehensive-research-agent`, `content-research-writer`, `cosmic-database`, `datacommons-client`, `documentation-lookup`, `drugbank-database`, `ena-database`, `ensembl-database`, `fda-database`, `geo-database`, `gwas-database`, `hmdb-database`, `hypothesis-generation`, `kegg-database`, `literature-matrix`, `literature-review`, `manuscript-as-code`, `market-research-reports`, `metabolomics-workbench-database`, `open-notebook`, `openalex-database`, `opentargets-database`, `paper-2-web`, `pdb-database`, `peer-review`, `pubchem-database`, `pubmed-database`, `pyzotero`, `reactome-database`, `research-grants`, `research-lookup`, `scholar-evaluation`, `scholarly-publishing`, `scientific-brainstorming`, `scientific-critical-thinking`, `scientific-reporting`, `scientific-writing`, `string-database`, `submission-checklist`, `uniprot-database`, `uspto-database`, `zinc-database`\n\n---\n\n### 🎨 Multimedia, Visualization & Documentation\n\n> **Making knowledge and data visible**: interactive chart generation, publication-quality scientific figures, slide creation, audio\u002Fvideo production, and deep read\u002Fwrite parsing of Word, PDF, and other office documents.\n\n`algorithmic-art`, `creating-data-visualizations`, `data-storytelling`, `datavis`, `doc`, `docs-review`, `docs-write`, `document-skills`, `docx`, `docx-comment-reply`, `figma`, `figma-implement-design`, `file-organizer`, `g2-legend-expert`, `generate-image`, `imagegen`, `infographics`, `latex-posters`, `latex-submission-pipeline`, `markdown-mermaid-writing`, `markitdown`, `matplotlib`, `pdf`, `plotly`, `pptx-posters`, `report-generator`, `scientific-schematics`, `scientific-slides`, `scientific-visualization`, `screenshot`, `seaborn`, `slides-as-code`, `smart-file-writer`, `speech`, `structured-content-storage`, `transcribe`, `venue-templates`, `video-studio`, `visualization-best-practices`, `vscode-release-notes-writer`, `writing-docs`\n\n---\n\n### 🔌 External Integrations, Automation & Deployment\n\n> **Breaking the limits of the runtime**: seamlessly connect external browsers, design platforms, and cloud services via MCP protocol and Playwright automation — plus CI\u002FCD pipeline support and one-click automated deployment.\n\n`aios-devops`, `alpha-vantage`, `claude-skills`, `commit-with-reflection`, `denario`, `digital-brain`, `edgartools`, `flashrag-evidence`, `fred-economic-data`, `geomaster`, `gh-address-comments`, `gh-fix-ci`, `hedgefundmonitor`, `hypogenic`, `iso-13485-certification`, `jupyter-notebook`, `knowledge-steward`, `mcp-integration`, `modal`, `modal-labs`, `netlify-deploy`, `openai-docs`, `perplexity-search`, `playwright`, `prowler-docs`, `scrapling`, `sentry`, `skypilot-multi-cloud-orchestration`, `vercel-deploy`\n\n\u003C\u002Fdetails>\n\n\u003Cbr\u002F>\n\n---\n\n\n## 📊 Why is it powerful?\n\n_Now for the numbers. This isn't a demo project — it's a running system._\n\nThe runtime core behind **VibeSkills** is **VCO**. This is not a single-point tool or a \"code completion\" script — it is a **super-capability network** that has been deeply integrated and governed:\n\n\u003Cbr\u002F>\n\n\u003Cdiv align=\"center\">\n\n|                              🧩 Skill Modules                               |                            🌍 Ecosystem                            |                               ⚖️ Governance Rules                                |\n| :---------------------------------------------------------------------: | :---------------------------------------------------------------: | :----------------------------------------------------------------------: |\n| \u003Ch2>340+\u003C\u002Fh2>Directly callable Skills\u003Cbr\u002F>covering the full chain from requirements to delivery | \u003Ch2>19+\u003C\u002Fh2>Absorbed high-value upstream\u003Cbr\u002F>open-source projects and best practices | \u003Ch2>129\u003C\u002Fh2>Policy rules and contracts\u003Cbr\u002F>ensuring stable, traceable, divergence-free execution |\n\n\u003C\u002Fdiv>\n\n\u003Cbr\u002F>\n\n---\n\n\n## ⚙️ Installation & Skills Management\n\n_Skills keep growing — but you don't need to manage them individually._\n\n### Uninstall: Owned-only cleanup\n\nRunning `uninstall.ps1` or `uninstall.sh --host \u003Chost>` is the partner surface to install. By default it performs a ledger-first, owned-only cleanup that only touches paths recorded in `.vibeskills\u002Finstall-ledger.json`, `*.host-closure.json`, or the documented legacy surfaces. The bundled runtime keeps only the executable contract; the full governance explainer lives in the canonical repo at [`docs\u002Funinstall-governance.md`](https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\u002Fblob\u002Fmain\u002Fdocs\u002Funinstall-governance.md).\n\nThe `.vibeskills` brand is now split into two layers on purpose:\n\n- host-sidecar: `\u003Ctarget-root>\u002F.vibeskills\u002Fhost-settings.json`, `host-closure.json`, `install-ledger.json`, `bin\u002F*`\n- workspace-sidecar: `\u003Cworkspace-root>\u002F.vibeskills\u002Fproject.json`, `.vibeskills\u002Fdocs\u002Frequirements\u002F*`, `.vibeskills\u002Fdocs\u002Fplans\u002F*`, `.vibeskills\u002Foutputs\u002Fruntime\u002Fvibe-sessions\u002F*`\n\nThis keeps host install state separate from governed workspace\u002Fruntime artifacts while preserving the existing relative runtime contract. Explicit `ArtifactRoot` overrides still work when operators need a different artifact location.\n\n### Install: One entry, two public versions\n\n\u003Cdiv align=\"center\">\n\n| | Single Public Entry |\n|:---:|:---|\n| **Install** | [⚡ Prompt-based install (recommended)](https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\u002Fblob\u002Fmain\u002Fdocs\u002Finstall\u002Fone-click-install-release-copy.en.md) |\n| **Public versions inside the same entry** | `Full Version + Customizable Governance` \u002F `Framework Only + Customizable Governance` |\n| **Result** | choose host + action + version in one place, then copy the matching prompt |\n\n\u003C\u002Fdiv>\n\n### Customize: Add your own skills\n\n→ [Custom workflow & skill onboarding guide](https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\u002Fblob\u002Fmain\u002Fdocs\u002Finstall\u002Fcustom-workflow-onboarding.en.md)\n\n## 📦 Standing on the Shoulders of Giants\n\n_These capabilities weren't built from scratch. VibeSkills' foundation is the continuous integration of the best open-source solutions into one governed system._\n\nWe know that building in isolation can't keep pace with the rapidly evolving AI landscape. The core strength of VibeSkills comes from continuously absorbing the most mature methods and architectures from the open-source community, and bringing them under a single unified governance and orchestration system.\n\n> 🙏 **Special Thanks & Acknowledgements**\n>\n> This project continuously integrates, absorbs, and governs the core strengths of the following excellent open-source projects:\n>\n> `superpower` · `claude-scientific-skills` · `get-shit-done` · `aios-core` · `OpenSpec` · `ralph-claude-code` · `SuperClaude_Framework` · `spec-kit` · `Agent-S` · `mem0` · `scrapling` · `claude-flow` · `serena` · `everything-claude-code` · `DeepAgent` and more\n>\n> _Thank you to all authors for your generous contributions — without these brilliant stars, VibeSkills would not exist. We have done our best to properly attribute and credit all absorbed repositories. If anything was missed, please open an Issue and we will correct it promptly._\n\n\u003Cbr\u002F>\n\n---\n\n\n## 🚀 Getting Started\n\n_You know what this is now. All it takes from here is one prompt:_\n\n> ⚠️ **Invocation note**: This project uses a **Skills format architecture**. Please invoke it through your host environment's Skills invocation method — **do not** run it as a standalone CLI program.\n\n\u003Cbr\u002F>\n\n\u003Cdiv align=\"center\">\n\n| Host Environment | Invocation | Example |\n|:---:|:---:|:---|\n| **Claude Code** | `\u002Fvibe` | `I want you to design a XXX \u002Fvibe` |\n| **Codex** | `$vibe` | `I want you to design a XXXX $vibe` |\n| **OpenCode** | `\u002Fvibe` | `Use the vibe skill to plan this change.` |\n| **OpenClaw** | Skills entry | Refer to the host docs |\n| **Cursor \u002F Windsurf** | Skills entry | Refer to each platform's Skills docs |\n\n\u003C\u002Fdiv>\n\n\u003Cbr\u002F>\n\n> 💡 **Tip**: To keep every message within the VibeSkills governed workflow, append `$vibe` or `\u002Fvibe` to each of your messages. A message without the invocation syntax is treated as a regular request outside the governed runtime.\n\n**Currently supported platforms**: `codex` (most complete governed path) · `claude-code` · `cursor` · `windsurf` (supported install-and-use path with runtime-adapter integration) · `openclaw` (preview runtime-core path) · `opencode` (preview adapter path)\n\n\u003Cbr\u002F>\n\n---\n\n\u003Cdetails>\n\u003Csummary>\u003Cb>📚 Documentation & Installation Guides (click to expand)\u003C\u002Fb>\u003C\u002Fsummary>\n\n\u003Cbr\u002F>\n\n**Understand the system**\n\n- 📖 [System architecture & philosophy](https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\u002Fblob\u002Fmain\u002Fdocs\u002Fquick-start.en.md)\n- 📜 [VibeSkills Manifesto](https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\u002Fblob\u002Fmain\u002Fdocs\u002Fmanifesto.en.md)\n\n**Installation & Configuration**\n\n- ⚡️ [Prompt-based install (recommended)](https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\u002Fblob\u002Fmain\u002Fdocs\u002Finstall\u002Fone-click-install-release-copy.en.md)\n- 🧩 [Custom workflow onboarding](https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\u002Fblob\u002Fmain\u002Fdocs\u002Finstall\u002Fcustom-workflow-onboarding.en.md)\n- 📄 [OpenClaw host notes](https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\u002Fblob\u002Fmain\u002Fdocs\u002Finstall\u002Fopenclaw-path.en.md)\n- 📄 [OpenCode host notes](https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\u002Fblob\u002Fmain\u002Fdocs\u002Finstall\u002Fopencode-path.en.md)\n- 📁 [Manual copy install (offline)](https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\u002Fblob\u002Fmain\u002Fdocs\u002Finstall\u002Fmanual-copy-install.en.md)\n- 🛠 [Advanced install command reference](https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\u002Fblob\u002Fmain\u002Fdocs\u002Finstall\u002Frecommended-full-path.en.md)\n- 🧊 [Cold start & other environments](https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\u002Fblob\u002Fmain\u002Fdocs\u002Fcold-start-install-paths.en.md)\n\n\u003C\u002Fdetails>\n\n\u003Cbr\u002F>\n\n\u003Cdiv align=\"center\">\n\n### 🤝 Join the Community · Build Together\n\nGive it a try! If you have questions, ideas, or suggestions, feel free to open an issue — I'll take every piece of feedback seriously and make improvements.\n\n\u003Cbr\u002F>\n\n**This project is fully open source. All contributions are welcome!**\n\nWhether it's fixing bugs, improving performance, adding features, or improving documentation — every PR is deeply appreciated.\n\n```\nFork → Modify → Pull Request → Merge ✅\n```\n\n\u003Cbr\u002F>\n\n> ⭐ If this project helps you, a **Star** is the greatest support you can give!\n> Your support is the enriched uranium that fuels this nuclear-powered donkey 🫏\n\n\u003Cbr\u002F>\n\nThank you to the **LinuxDo** community for your support!\n\n[![LinuxDo](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FCommunity-LinuxDo-blue?style=for-the-badge)](https:\u002F\u002Flinux.do\u002F)\n\nTech discussions, AI frontiers, AI experience sharing — all at Linuxdo!\n\n\u003C\u002Fdiv>\n\n\u003Cbr\u002F>\n\n---\n\n## Star History\n\u003Cdiv align=\"center\">\n\u003Ca href=\"https:\u002F\u002Fwww.star-history.com\u002F?repos=foryourhealth111-pixel%2FVibe-Skills&type=date&legend=top-left\">\n \u003Cpicture>\n   \u003Csource media=\"(prefers-color-scheme: dark)\" srcset=\"https:\u002F\u002Fapi.star-history.com\u002Fimage?repos=foryourhealth111-pixel\u002FVibe-Skills&type=date&theme=dark&legend=top-left\" \u002F>\n   \u003Csource media=\"(prefers-color-scheme: light)\" srcset=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fforyourhealth111-pixel_Vibe-Skills_readme_c7193a8d707c.png\" \u002F>\n   \u003Cimg alt=\"Star History Chart\" src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fforyourhealth111-pixel_Vibe-Skills_readme_c7193a8d707c.png\" \u002F>\n \u003C\u002Fpicture>\n\u003C\u002Fa>\n\n---\n\n\u003Cdiv align=\"center\">\n  \u003Cp>\u003Ci>Transform the parts of real work most prone to going off the rails into a system that is more callable, more governable, and more maintainable over time.\u003C\u002Fi>\u003C\u002Fp>\n  \u003Cbr\u002F>\n  \u003Csub>Made with ❤️  ·  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\">GitHub\u003C\u002Fa>  ·  \u003Ca href=\".\u002FREADME.zh.md\">中文\u003C\u002Fa>\u003C\u002Fsub>\n\u003C\u002Fdiv>\n","\u003Cdiv align=\"right\">\n  \u003Cb>🇬🇧 English\u003C\u002Fb>  |  \u003Ca href=\".\u002FREADME.zh.md\">🇨🇳 中文\u003C\u002Fa>\n\u003C\u002Fdiv>\n\n\u003Cbr\u002F>\n\n\u003Cdiv align=\"center\">\n\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\">\n  \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fforyourhealth111-pixel_Vibe-Skills_readme_f9e121099867.png\" alt=\"VibeSkills Typing Logo\" \u002F>\n\u003C\u002Fa>\n\n\u003Cbr\u002F>\n\n\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fforyourhealth111-pixel_Vibe-Skills_readme_947a6dc2a01f.png\" width=\"260px\" alt=\"VibeSkills Logo\"\u002F>\n\n\u003Cbr\u002F>\u003Cbr\u002F>\n\n### 不止于技能集合——你的**个人AI操作系统**\n\n只要你的AI支持技能，VibeSkills就能运行。涵盖编码、研究、数据科学和创意工作的340多种技能。\n\n\u003Cbr\u002F>\n\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\u002Fstargazers\">\n  \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fforyourhealth111-pixel\u002FVibe-Skills?style=for-the-badge&logo=github&color=7B61FF&label=STARS\" alt=\"stars\">\n\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\u002Fnetwork\u002Fmembers\">\n  \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fforks\u002Fforyourhealth111-pixel\u002FVibe-Skills?style=for-the-badge&logo=git&color=45a1ff&label=FORKS\" alt=\"forks\">\n\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\u002Fpulse\">\n  \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Flast-commit\u002Fforyourhealth111-pixel\u002FVibe-Skills?style=for-the-badge&logo=git-lfs&color=32CD32&label=MOMENTUM\" alt=\"last commit\">\n\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgitcgr.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\">\n  \u003Cimg src=\"https:\u002F\u002Fgitcgr.com\u002Fbadge\u002Fforyourhealth111-pixel\u002FVibe-Skills.svg\" alt=\"gitcgr\" \u002F>\n\u003C\u002Fa>\n\n\u003Cbr\u002F>\u003Cbr\u002F>\n\n\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fforyourhealth111-pixel_Vibe-Skills_readme_510b79be381e.png\" alt=\"Visitors\">\n \n\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FArchitecture-VCO_Runtime-orange?style=for-the-badge\" alt=\"Arch\">\n \n\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FSkills-340%2B-blueviolet?style=for-the-badge\" alt=\"Skills Count\">\n\n\u003Cbr\u002F>\u003Cbr\u002F>\n\n🧠 规划 · 🛠️ 工程 · 🤖 AI · 🔬 研究 · 🎨 创作\n\n\u003Cbr\u002F>\n\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\u002Fblob\u002Fmain\u002Fdocs\u002Finstall\u002Fone-click-install-release-copy.en.md\">\n  \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002F⚡_Get_Started-7B61FF?style=for-the-badge\" alt=\"Install\">\n\u003C\u002Fa>\n \n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\u002Fblob\u002Fmain\u002Fdocs\u002Fquick-start.en.md\">\n  \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002F📖_Quick_Start-2d3748?style=for-the-badge\" alt=\"Docs\">\n\u003C\u002Fa>\n \n\u003Ca href=\".\u002FREADME.zh.md\">\n  \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002F🇨🇳_中文-45a1ff?style=for-the-badge\" alt=\"Chinese\">\n\u003C\u002Fa>\n\n\u003Cbr\u002F>\u003Cbr\u002F>\n\n\u003Ckbd>安装\u003C\u002Fkbd>  → \n\u003Ckbd>\u002Fvibe 或 $vibe\u003C\u002Fkbd>  → \n\u003Ckbd>智能路由\u003C\u002Fkbd>  → \n\u003Ckbd>M \u002F L \u002F XL 执行\u003C\u002Fkbd>  → \n\u003Ckbd>治理验证\u003C\u002Fkbd>  → \n\u003Ckbd>✅ 交付\u003C\u002Fkbd>\n\n\u003C\u002Fdiv>\n\n## 📋 目录\n\n- [它有何不同](#-what-makes-it-different)\n- [适合哪些人](#-who-is-it-for)\n- [智能路由](#-intelligent-routing-how-340-skills-collaborate-without-conflict)\n- [记忆系统](#-memory-system-ai-that-truly-remembers)\n- [完整能力图](#-full-capability-map-your-all-in-one-workbench)\n- [安装与管理](#️-installation--skills-management)\n- [开始使用](#-getting-started)\n\n\n\u003Cdetails>\n\u003Csummary>\u003Cb>🔑 刚来？关键术语速览（点击展开）\u003C\u002Fb>\u003C\u002Fsummary>\n\n\u003Cbr\u002F>\n\n| 术语 | 白话解释 |\n|:---|:---|\n| **VibeSkills \u002F VCO** | 这个项目。VCO = Vibe Code Orchestrator — 技能背后的运行时引擎。 |\n| **技能** | 一个专注的能力模块（例如 `tdd-guide`、`code-review`）。可以将技能视为系统按需调用的专家助手。 |\n| **受治理的运行时** | 当你调用 `\u002Fvibe` 时，系统会遵循一套结构化流程——明确 → 计划 → 执行 → 验证——而不是盲目行动。 |\n| **规范路由器** | 内部逻辑，用于决定针对你的任务激活哪个技能。只需调用 `\u002Fvibe`，它就会自动路由。 |\n| **M \u002F L \u002F XL 等级** | 任务复杂度级别。M = 快速聚焦任务，L = 多步骤任务，XL = 包含并行工作的大型任务。自动选择。 |\n| **冻结需求** | 一旦你确认计划，它就会被“冻结”——系统不会在任务中途悄悄更改范围。 |\n| **根车道 \u002F 子车道** | 在XL任务中，有一个“根”协调员和“子”工作者代理。防止并行代理产生冲突结果。 |\n| **证明包** | 任务确实正确完成的证据——测试结果、输出、验证日志。 |\n\n\u003C\u002Fdetails>\n\n> [!IMPORTANT]\n> ### 🎯 核心愿景\n>\n> VibeSkills 随时代不断进化——确保其始终保持真正的实用性，同时**大幅降低前沿振动编码技术的门槛**，消除新AI工具带来的认知焦虑和陡峭的学习曲线。\n>\n> **无论你是否有编程背景，都可以以最小的努力直接利用最先进的AI能力。** \n> AI带来的生产力提升应该惠及所有人。\n\n\u003Cbr\u002F>\n\n---\n\n## ✨ 它有何不同？\n\n> 传统技能仓库的回答是：_“我有哪些工具？”_\n> **VibeSkills直击重度AI用户的核心痛点：_“我该如何管理和调用大量技能，高效可靠地完成工作？”_**\n\n\u003Csub>兼容**Claude Code** · **Codex** · **Windsurf** · **OpenClaw** · **OpenCode** · **Cursor**，以及任何支持Skills协议的AI环境。原生兼容**MCP**。\u003C\u002Fsub>\n\n\u003Cbr\u002F>\n\n\u003Cdiv align=\"center\">\n\n| ❌  传统痛点（你可能深有体会） | ✅  VibeSkills解决方案（我们打造的功能） |\n|:---|:---|\n| **技能从未被激活**：仓库中有数百种能力，但AI很少记得去使用它们——激活率极低。 | **🧠 智能路由**：系统会根据上下文自动路由到合适的技能，无需记忆技能列表。 |\n| **盲目执行**：AI不澄清需求就直接动手——速度快却偏离目标，项目逐渐变成黑箱。 | **🧭 受控的工作流**：统一流程中强制执行“澄清→验证→追溯”；每一步都可审计。 |\n| **工具冲突**：插件与工作流之间缺乏协调，导致环境污染或陷入无限循环。 | **🧩 全局治理**：129条合约规则定义了安全边界和回退机制，确保长期稳定性。 |\n| **混乱的工作空间**：长时间使用后，仓库变得杂乱无章；新接手的Agent会错过项目细节，造成交接断层。 | **📁 语义目录治理**：采用固定架构的文件存储方式，使任何新的AI对话都能立即理解项目背景。 |\n| **AI不良习惯**：清理备份时误删主文件；编写无声回退逻辑后却自信宣称“已完成”。 | **🛡️ 内置安全规则**：受控执行默认阻止危险的大规模删除和盲目的递归清除操作；回退机制必须始终显示明确警告。 |\n| **手动维护工作流纪律**：用户需凭经验自行维护AI协作流程——学习成本高。 | **🚦 框架引导的端到端流程**：需求→计划→多智能体执行→自动化测试迭代——全程托管。 |\n| **多智能体运行中的技能调度混乱**：难以为不同任务分配恰当的技能给每个智能体。 | **🤖 自动技能调度**：多智能体工作流会自动将相应技能分配给每个智能体的任务。 |\n\n\u003C\u002Fdiv>\n\n\u003Cbr\u002F>\n\n---\n\n\n## 👥 适合哪些人？\n\n_以上哪些痛点戳中了你？找到自己的位置——接下来的内容会让你更有共鸣。_\n\n\u003Cdetails>\n\u003Csummary>这是否适合你？点击展开\u003C\u002Fsummary>\n\n\u003Cbr\u002F>\n\n\u003Cdiv align=\"center\">\n\n| 目标用户 | 描述 |\n|:---:|:---|\n| 🎯 **需要可靠交付的用户** | 希望AI成为可靠的伙伴，而非脱缰野马 |\n| ⚡ **重度依赖AI\u002FAgent的高级用户** | 需要一个统一的基础来支撑大规模工作流 |\n| 🏢 **对标准化要求较高的小型团队** | 希望AI工作流更易维护且可交接 |\n| 😩 **因技能泛滥而疲惫的从业者** | 已经厌倦了到处寻找工具——只想得到开箱即用的解决方案 |\n\n\u003C\u002Fdiv>\n\n> _如果你只是想找一个小小的脚本，那这可能有些大材小用了。但如果你想更可靠、顺畅、可持续地使用AI——这就是你不可或缺的基础。_\n\n\u003C\u002Fdetails>\n\n\u003Cbr\u002F>\n\n---\n\n\n## 🔀 智能路由：340+技能如何协作而不冲突\n\n_你已经知道它适合你。下一个问题：一个系统里有340多种技能——它们如何互不干扰呢？_\n\n面对340多种技能，你可能会想：_“相似的技能会不会发生冲突？系统怎么知道该用哪一个？”_\n\n### 路由的工作原理\n\nVibeSkills采用**规范路由器**作为唯一的权威路由决策中心：\n\n```mermaid\ngraph LR\n    A[用户任务] --> B{规范路由器}\n    B --> C[意图识别]\n    C --> D[关键词提取]\n    D --> E[技能匹配]\n    E --> F[冲突检测]\n    F --> G[优先级排序]\n    G --> H[路由决策]\n    H --> I[执行技能]\n\n    style B fill:#7B61FF,stroke:#fff,stroke-width:2px,color:#fff\n    style F fill:#FF9800,stroke:#fff,stroke-width:2px,color:#fff\n```\n\nVibeSkills遵循“澄清→规划→执行→验证”的受控工作流，确保每个任务都经过完整的质量控制：\n\n- **需求澄清**：像`speckit-clarify`这样的技能会定义清晰的边界和验收标准。\n- **架构规划**：像`aios-architect`这样的技能会设计实现路径。\n- **执行层**：按需调用340多种技能来完成实际工作。\n- **质量验证**：像`tdd-guide`和`code-review`这样的技能会确保交付质量。\n\n---\n\n### 为什么采用这种设计？\n\n传统的技能仓库允许AI“自由选择”——结果是：\n\n- ❌ AI记不住都有哪些技能。\n- ❌ 相似技能之间会发生冲突。\n- ❌ 执行路径不可预测。\n\nVibeSkills的路由机制则保证了：\n\n- ✅ **确定性**：同一任务总是遵循相同的路由逻辑。\n- ✅ **可追溯性**：每一次路由决策都有明确的理由。\n- ✅ **可控性**：用户可以通过显式调用（如`\u002Fvibe`）覆盖默认路由。\n- ✅ **稳定性**：129条治理规则可以防止冲突和偏差。\n\n---\n\n### M \u002F L \u002F XL 执行级别\n\n在选择主要技能后，路由器还会根据任务复杂度自动确定执行级别：\n\n\u003Cdiv align=\"center\">\n\n| 级别 | 使用场景 | 特点 |\n|:---:|:---|:---|\n| **M** | 范围狭窄、边界清晰的工作 | 单代理、令牌高效、响应迅速 |\n| **L** | 中等复杂度，需要设计、规划和审查 | 按照计划步骤进行原生串行执行；仅在明确计划时才委托有限的子单元 |\n| **XL** | 大型任务——可并行化、长时间运行、多代理分波执行 | 波次顺序编排，仅对独立单元实现步骤级的有限并行 |\n\n\u003C\u002Fdiv>\n\n> 系统会在需求澄清之后、计划执行之前自动选择级别。用户只需调用 `\u002Fvibe` 或 `$vibe` 即可。\n>\n> 当系统内部调用专业技能（如 `tdd-guide` 或 `code-review`）时，其范围始终限定于特定阶段——它们提供协助，但不接管整体协调。在 XL 级别的多代理任务中，工作者代理（子通道）可以建议寻求专业帮助，但必须由协调器（根节点）批准后方可执行。\n>\n> 你也可以明确表达偏好：\n> ```text\n> 请按照计划执行此任务，启动 XL 级工作流 \u002Fvibe\n> ```\n\n---\n\n\u003Cdetails>\n\u003Csummary>\u003Cb>🔍 路由常见问题解答（点击展开）\u003C\u002Fb>\u003C\u002Fsummary>\n\n\u003Cbr\u002F>\n\n**每个任务只有一条路由还是多条？**\n\n核心原则：一个任务通常会路由到一项主要技能，但该技能可以在执行过程中调用其他技能作为子流程。\n\n- **单一主路由**：标准路由器会选择**最匹配的主要技能**\n- **技能组合**：主要技能可在执行过程中根据需要调用其他技能（例如，`vibe` 可以调用 `speckit-clarify`、`aios-architect` 等）\n- **受控协调**：多技能协作由治理规则控制，而非随意组合\n\n\u003Cbr\u002F>\n\n**相似技能之间的冲突如何处理？**\n\n当多个技能看起来都能完成任务时，路由器通过以下方式避免冲突：\n\n1. **优先级规则**：每项技能都有明确的优先级和适用场景\n2. **上下文匹配**：分析任务复杂度、多阶段需求以及用户的明确偏好\n3. **互斥规则**：129 条规则包括防止冲突组合的排除规则\n4. **优雅降级**：当首选技能不可用时，按优先级回退——不会出现无限循环\n\n\u003Cbr\u002F>\n\n**选项过多是否会导致令牌消耗激增？**\n\n不会。路由并不会将所有选项一股脑地输入模型——它采用智能触发机制：\n\n```\n用户指令 → AI 辅助治理提取意图关键词 → 关键词触发技能路由\n```\n\n治理框架会增加约 3 万的初始上下文开销，但不会导致令牌爆炸式增长。\n\n\u003Cbr\u002F>\n\n**真实案例：用户说“帮我重构这个项目”**\n\n1. 意图识别 → 复杂的重构任务\n2. 关键词提取 → 重构、项目、代码质量\n3. 技能匹配 → `vibe` \u002F `autonomous-builder` \u002F `systematic-debugging`\n4. 路由决策 → 选择 `vibe`（重构需要多阶段：澄清 → 计划 → 执行 → 验证）\n\n\u003C\u002Fdetails>\n\n\u003Cbr\u002F>\n\n---\n\n\n## 🧠 记忆系统：真正能记住的 AI\n\n_路由解决了“该用哪项技能”的问题。但还有一个更深层次的问题：对话结束后，AI 还记得你吗？_\n\n是不是很熟悉？\n\n\u003Cdiv align=\"center\">\n\n| ❌ 痛点 | ✅ VibeSkills 解决方案 | 组件 |\n|:---|:---|:---:|\n| 每次新会话都要重新解释项目背景 | 架构决策和规范在启动时自动加载 | `Serena` |\n| AI 再次遇到同样的 bug；见解随上下文消失 | 一句话即可永久保存至 Obsidian + GitHub | `knowledge-steward` |\n| 长时间任务——AI 逐渐“忘记”早期上下文 | 会话内语义向量缓存，可即时检索 | `ruflo` |\n| 跨项目知识无法积累 | 实体关系图会随着时间不断丰富 | `Cognee` |\n| 长时间任务被打断——难以交接给新代理 | 自动折叠为工作、工具和证据记忆 | `deepagent-memory-fold` |\n\n\u003C\u002Fdiv>\n\n\u003Cbr\u002F>\n\n\u003Cdetails>\n\u003Csummary>\u003Cb>📐 展开：四层架构、记忆技能与治理规则\u003C\u002Fb>\u003C\u002Fsummary>\n\n\u003Cbr\u002F>\n\nVibeSkills 构建了一个**四层记忆系统**——针对每种记忆需求都有一套权威组件：\n\n| 层级 | 组件 | 范围 | 核心目的 |\n|:---:|:---:|:---:|:---|\n| **L1 会话** | `state_store` | 当前会话 | 执行进度、中间结果、临时状态——始终在线的“工作台” |\n| **L2 项目** | `Serena` | 当前项目 | 架构决策、规范——仅在用户明确确认后才会写入 |\n| **L3 短期语义** | `ruflo` | 会话内 | 长时间任务中用于快速检索上下文的向量缓存 |\n| **L4 长期图谱** | `Cognee` | 跨会话 | 实体链接、关系图、长期知识积累 |\n\n> **可选扩展**：`mem0` 作为个人偏好后端（需主动启用）；`Letta` 提供内存块映射词汇表——两者均不取代这四个标准层级。\n\n\u003Cbr\u002F>\n\n**三项专用记忆技能**\n\n| 技能 | 角色 | 触发条件 |\n|:---:|:---|:---|\n| `knowledge-steward` | **知识守护者**：将见解、bug 修复和提示永久保存到 Obsidian + GitHub | “保存这个提示” \u002F “记录这个 bug” \u002F “保存这个见解” |\n| `digital-brain` | **第二大脑**：结构化的个人知识库——身份、内容、网络、回顾 | 直接调用；非常适合用作个人知识操作系统 |\n| `deepagent-memory-fold` | **上下文折叠**：将大量上下文压缩成结构化的工作\u002F工具\u002F证据记忆，以便无缝交接 | 在上下文达到上限时或手动触发 |\n\n\u003Cbr\u002F>\n\n**治理**：单一事实来源（无双轨制）· 明确的写入权限（`Serena` 需用户确认）· 永久禁用 `episodic-memory` · `mem0` 仅限于个人偏好 · 每个外部后端都设有紧急关闭开关\n\n\u003C\u002Fdetails>\n\n\n---\n\n## ✦ 全能力地图：您的全能工作台\n\n_路由与记忆构成了调度神经系统。以下是它们驱动的端到端全能力链。_\n\n在“真实工作流”中展开，VibeSkills 已经布局了一条完整的 **端到端能力链**：\n\n\u003Cbr\u002F>\n\n\u003Cdiv align=\"center\">\n\n| 领域 | 覆盖范围 | 代表性引擎 |\n|:---|:---|:---|\n| **💡 需求与澄清** | 再也不用黑箱启动：将模糊想法转化为边界清晰、可验证的问题定义 | `brainstorming`、`speckit-clarify` |\n| **📋 规划与拆解** | 将宏伟目标分解为规格、计划、任务、里程碑和执行流程 | `writing-plans`、`speckit-specify`、`aios-po` |\n| **🏗️ 架构与技术选型** | 设计前端\u002F后端边界、API、数据层、部署方案以及技术栈对比 | `aios-architect`、`architecture-patterns` |\n| **💻 开发与实现** | 新功能开发、脚手架搭建、工程集成以及跨文件的精准实现 | `autonomous-builder`、`speckit-implement` |\n| **🔧 调试与重构** | 不止于表面修复：定位错误、分析根因，恢复项目级可维护性 | `error-resolver`、`systematic-debugging` |\n| **🛡️ 测试与质量控制** | 单元测试、回归验证、质量门——完成前必须通过的强制性验证 | `tdd-guide`、`aios-qa`、`code-review` |\n| **🚀 协作与发布** | 处理 Issue\u002FPR、修复 CI、评审流程及自动化部署 | `aios-devops`、`gh-fix-ci`、`vercel-deploy` |\n| **🤖 复合工作流** | 冻结需求、分派任务、多 Agent 协调、执行追踪、环境清理 | `vibe`、`swarm_*`、`hive-mind-advanced` |\n| **🔌 外部生态** | 桥接浏览器、网页爬取、设计文件、第三方服务以及上下文记忆 | `mcp-integration`、`playwright`、`scrapling` |\n| **📊 数据与 AI 工程** | EDA、清洗与统计分析，直至模型训练、RAG 检索和实验跟踪 | `senior-ml-engineer`、`statistical-analysis` |\n| **🔬 研究与生命科学** | **核心实力**：文献综述、生物信息学、单细胞分析、药物发现 | `literature-review`、`biopython`、`scanpy` |\n| **📐 数学与科学计算** | 符号推导、贝叶斯建模、多目标优化、仿真以及量子计算 | `sympy`、`pymc-bayesian-modeling`、`qiskit` |\n| **🎨 多媒体与演示** | 交互式图表、出版级图表、幻灯片制作以及音视频制作 | `plotly`、`generate-image`、`video-studio` |\n\n\u003C\u002Fdiv>\n\n\u003Cbr\u002F>\n\n\u003Cdetails>\n\u003Csummary>\u003Cb>👉 展开：探索完整的 340+ 全栈能力矩阵\u003C\u002Fb>\u003C\u002Fsummary>\n\n\u003Cbr\u002F>\n\n> 💡 **为什么治理很重要**：下方庞大的技能库并非孤立脚本的集合——它是由 VCO 运行时治理的生态系统。通过领域矩阵分类，系统会在合适的上下文节点自动调用正确的工具，无需您手动搜索技能。\n\n---\n\n### 🧠 需求、规划与产品管理\n\n> **将宏大想法转化为可执行计划**：需求洞察、问题定义、Sprint 计划、任务拆解以及约束收集。确保方向明确、边界清晰、里程碑可验证，再开始编写一行代码。\n\n`.system`、`aios-pm`、`aios-po`、`aios-sm`、`aios-squad-creator`、`aios-ux-design-expert`、`brainstorming`、`create-plan`、`designing-experiments`、`planning-with-files`、`shared-templates`、`speckit-analyze`、`speckit-checklist`、`speckit-clarify`、`speckit-constitution`、`speckit-plan`、`speckit-specify`、`speckit-tasks`、`speckit-taskstoissues`、`subagent-driven-development`、`think-harder`、`treatment-plans`、`ux-researcher-designer`、`writing-plans`\n\n---\n\n### 🛠️ 软件工程与架构\n\n> **真正的工程基础**：从脚手架搭建、跨文件修改、API 设计到微服务架构评估。不仅仅是代码输出——还包括上下文记忆、工具链编排以及多阶段智能 Agent 协调。\n\n`aios-architect`、`aios-dev`、`aios-master`、`architecture-patterns`、`autonomous-builder`、`cancel-ralph`、`coding-tutor`、`context-fundamentals`、`context-hunter`、`cs-foundations`、`deepagent-memory-fold`、`deepagent-toolchain-plan`、`evaluating-code-models`、`get-available-resources`、`hive-mind-advanced`、`local-vco-roles`、`nowait-reasoning-optimizer`、`prompt-lookup`、`ralph-loop`、`skill-creator`、`skill-lookup`、`spec-kit-vibe-compat`、`speckit-implement`、`superclaude-framework-compat`、`theme-factory`、`vibe`、`webthinker-deep-research`\n\n---\n\n### 🔧 调试、测试与质量保证\n\n> **守护代码与系统的生命线**：单元测试、根因分析、依赖冲突解决、安全漏洞审查以及完整的 TDD 指南——确保系统不会陷入“每次改动就出错”的黑箱状态。\n\n`aios-qa`、`build-error-resolver`、`code-review`、`code-review-excellence`、`code-reviewer`、`data-quality-checker`、`data-quality-frameworks`、`debugging-strategies`、`deslop`、`detecting-performance-regressions`、`error-resolver`、`evals-context`、`experiment-failure-analysis`、`generating-test-reports`、`ml-data-leakage-guard`、`performance-testing`、`property-based-testing`、`providing-performance-optimization-advice`、`receiving-code-review`、`requesting-code-review`、`reviewing-code`、`security-best-practices`、`security-ownership-map`、`security-reviewer`、`security-threat-model`、`systematic-debugging`、`tdd-guide`、`verification-before-completion`、`verification-quality-assurance`、`windows-hook-debugging`\n\n---\n\n### 📊 数据分析与统计建模\n\n> **让数据说出真相**：一站式数据处理引擎，涵盖数据清洗、缺失值处理、探索性数据分析（EDA），直至高级统计检验、回归模型和时间序列预测。\n\n`aios-data-engineer`、`anomaly-detector`、`correlation-analyzer`、`dask`、`data-artist`、`data-exploration-visualization`、`data-normalization-tool`、`detecting-data-anomalies`、`excel-analysis`、`exploratory-data-analysis`、`feature-importance-analyzer`、`geopandas`、`hypothesis-testing`、`metric-calculator`、`networkx`、`performing-causal-analysis`、`performing-regression-analysis`、`polars`、`preprocessing-data-with-automated-pipelines`、`regression-analysis-helper`、`running-clustering-algorithms`、`scientific-data-preprocessing`、`splitting-datasets`、`spreadsheet`、`statistical-analysis`、`statistics-math`、`statsmodels`、`usfiscaldata`、`vaex`、`xlsx`\n\n---\n\n### 🤖 机器学习与人工智能工程\n\n> **全栈式AI模型开发**：不仅仅是调用API——还包括特征工程、模型训练、微调、可解释性（SHAP）、大模型评估（Evals）以及强化学习工作流。\n\n`LQF_Machine_Learning_Expert_Guide`, `aeon`, `datamol`, `deepchem`, `embedding-strategies`, `engineering-features-for-machine-learning`, `evaluating-llms-harness`, `evaluating-machine-learning-models`, `explaining-machine-learning-models`, `geniml`, `ml-pipeline-workflow`, `openai-knowledge`, `pufferlib`, `pytorch-lightning`, `scikit-learn`, `scikit-survival`, `senior-computer-vision`, `senior-data-scientist`, `senior-ml-engineer`, `senior-prompt-engineer`, `shap`, `similarity-search-patterns`, `sparse-autoencoder-training`, `stable-baselines3`, `tensorboard`, `timesfm-forecasting`, `torch-geometric`, `torch_geometric`, `torchdrug`, `training-machine-learning-models`, `transformer-lens-interpretability`, `transformers`, `umap-learn`, `unsloth`, `weights-and-biases`\n\n---\n\n### 🧬 生命科学与生物信息学\n\n> **强大的跨学科综合平台**：单细胞测序分析、蛋白质结构折叠、药物分子发现、基因组比对——与基于云的生物学实验室系统无缝集成。\n\n`adaptyv`, `alphafold-database`, `anndata`, `arboreto`, `benchling-integration`, `biopython`, `bioservices`, `cellxgene-census`, `cobrapy`, `deeptools`, `diffdock`, `dnanexus-integration`, `esm`, `etetoolkit`, `flowio`, `gene-database`, `gget`, `ginkgo-cloud-lab`, `gtars`, `histolab`, `imaging-data-commons`, `labarchive-integration`, `lamindb`, `latchbio-integration`, `matchms`, `medchem`, `molfeat`, `neurokit2`, `neuropixels-analysis`, `omero-integration`, `opentrons-integration`, `pathml`, `protocolsio-integration`, `pydeseq2`, `pydicom`, `pyhealth`, `pylabrobot`, `pyopenms`, `pysam`, `pytdc`, `rdkit`, `scanpy`, `scikit-bio`, `scvi-tools`, `tiledbvcf`\n\n---\n\n### 🔬 科学计算与数学逻辑\n\n> **精确推导与复杂系统仿真**：符号数学、贝叶斯概率编程、量子计算仿真、多目标优化，以及严格的命题逻辑和数学证明辅助工具。\n\n`astropy`, `cirq`, `dialectic`, `fluidsim`, `gradient-methods`, `math`, `math-model-selector`, `math-tools`, `mathematical-logic-expert`, `matlab`, `pennylane`, `pymatgen`, `pymc`, `pymc-bayesian-modeling`, `pymoo`, `propositional-logic`, `qiskit`, `qutip`, `rowan`, `simpy`, `sympy`, `xan`\n\n---\n\n### 📚 科学文献与学术写作\n\n> **提升学术生产力的快速通道**：可在数十个数据库中进行精准检索（PubMed、arXiv等），系统化整理综述矩阵，管理引用文献，并完成从草稿撰写到同行评审的完整发表流程。\n\n`bgpt-paper-search`, `biorxiv-database`, `brenda-database`, `chembl-database`, `citation-management`, `clinical-decision-support`, `clinical-reports`, `clinicaltrials-database`, `clinpgx-database`, `clinvar-database`, `comprehensive-research-agent`, `content-research-writer`, `cosmic-database`, `datacommons-client`, `documentation-lookup`, `drugbank-database`, `ena-database`, `ensembl-database`, `fda-database`, `geo-database`, `gwas-database`, `hmdb-database`, `hypothesis-generation`, `kegg-database`, `literature-matrix`, `literature-review`, `manuscript-as-code`, `market-research-reports`, `metabolomics-workbench-database`, `open-notebook`, `openalex-database`, `opentargets-database`, `paper-2-web`, `pdb-database`, `peer-review`, `pubchem-database`, `pubmed-database`, `pyzotero`, `reactome-database`, `research-grants`, `research-lookup`, `scholar-evaluation`, `scholarly-publishing`, `scientific-brainstorming`, `scientific-critical-thinking`, `scientific-reporting`, `scientific-writing`, `string-database`, `submission-checklist`, `uniprot-database`, `uspto-database`, `zinc-database`\n\n---\n\n### 🎨 多媒体、可视化与文档编写\n\n> **让知识与数据可视化**：交互式图表生成、出版级科学图形绘制、幻灯片制作、音视频制作，以及对Word、PDF等办公文档的深度读写解析。\n\n`algorithmic-art`, `creating-data-visualizations`, `data-storytelling`, `datavis`, `doc`, `docs-review`, `docs-write`, `document-skills`, `docx`, `docx-comment-reply`, `figma`, `figma-implement-design`, `file-organizer`, `g2-legend-expert`, `generate-image`, `imagegen`, `infographics`, `latex-posters`, `latex-submission-pipeline`, `markdown-mermaid-writing`, `markitdown`, `matplotlib`, `pdf`, `plotly`, `pptx-posters`, `report-generator`, `scientific-schematics`, `scientific-slides`, `scientific-visualization`, `screenshot`, `seaborn`, `slides-as-code`, `smart-file-writer`, `speech`, `structured-content-storage`, `transcribe`, `venue-templates`, `video-studio`, `visualization-best-practices`, `vscode-release-notes-writer`, `writing-docs`\n\n---\n\n### 🔌 外部集成、自动化与部署\n\n> **突破运行时的限制**：通过MCP协议和Playwright自动化工具，无缝连接外部浏览器、设计平台和云服务——同时支持CI\u002FCD流水线和一键自动化部署。\n\n`aios-devops`, `alpha-vantage`, `claude-skills`, `commit-with-reflection`, `denario`, `digital-brain`, `edgartools`, `flashrag-evidence`, `fred-economic-data`, `geomaster`, `gh-address-comments`, `gh-fix-ci`, `hedgefundmonitor`, `hypogenic`, `iso-13485-certification`, `jupyter-notebook`, `knowledge-steward`, `mcp-integration`, `modal`, `modal-labs`, `netlify-deploy`, `openai-docs`, `perplexity-search`, `playwright`, `prowler-docs`, `scrapling`, `sentry`, `skypilot-multi-cloud-orchestration`, `vercel-deploy`\n\n\u003C\u002Fdetails>\n\n\u003Cbr\u002F>\n\n---\n\n\n## 📊 为什么它如此强大？\n\n_现在来看数据。这并不是一个演示项目——而是一个正在运行的系统。_\n\n**VibeSkills**背后的运行时核心是**VCO**。这并非单一工具或“代码补全”脚本——而是一个经过深度集成与治理的**超级能力网络**：\n\n\u003Cbr\u002F>\n\n\u003Cdiv align=\"center\">\n\n|                              🧩 技能模块                               |                            🌍 生态系统                            |                               ⚖️ 治理规则                                |\n| :---------------------------------------------------------------------: | :---------------------------------------------------------------: | :----------------------------------------------------------------------: |\n| \u003Ch2>340+\u003C\u002Fh2>直接可调用的技能\u003Cbr\u002F>覆盖从需求到交付的完整链条 | \u003Ch2>19+\u003C\u002Fh2>吸收了高价值的上游\u003Cbr\u002F>开源项目和最佳实践 | \u003Ch2>129\u003C\u002Fh2>政策规则和契约\u003Cbr\u002F>确保稳定、可追溯且无偏差的执行 |\n\n\u003C\u002Fdiv>\n\n\u003Cbr\u002F>\n\n---\n\n\n## ⚙️ 安装与技能管理\n\n_技能在不断增长——但你无需单独管理它们。_\n\n### 卸载：仅清理自有资源\n\n运行 `uninstall.ps1` 或 `uninstall.sh --host \u003Chost>` 是安装的对应卸载接口。默认情况下，它会执行以账本优先、仅清理自有资源的操作，只触及记录在 `.vibeskills\u002Finstall-ledger.json`、`*.host-closure.json` 文件中，或文档中列出的旧版接口上的路径。打包的运行时仅保留可执行契约；完整的治理说明位于规范仓库中的 [`docs\u002Funinstall-governance.md`](https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\u002Fblob\u002Fmain\u002Fdocs\u002Funinstall-governance.md)。\n\n`.vibeskills` 品牌现在被有意划分为两层：\n\n- 主机侧边车：`\u003Ctarget-root>\u002F.vibeskills\u002Fhost-settings.json`、`host-closure.json`、`install-ledger.json`、`bin\u002F*`\n- 工作区侧边车：`\u003Cworkspace-root>\u002F.vibeskills\u002Fproject.json`、`.vibeskills\u002Fdocs\u002Frequirements\u002F*`、`.vibeskills\u002Fdocs\u002Fplans\u002F*`、`.vibeskills\u002Foutputs\u002Fruntime\u002Fvibe-sessions\u002F*`\n\n这样做可以将主机端的安装状态与受治理的工作区\u002F运行时工件分开，同时保留现有的相对运行时契约。当运维人员需要不同的工件存放位置时，显式指定的 `ArtifactRoot` 仍然有效。\n\n### 安装：一个入口，两个公开版本\n\n\u003Cdiv align=\"center\">\n\n| | 单一公共入口 |\n|:---:|:---|\n| **安装** | [⚡ 基于提示的安装（推荐）](https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\u002Fblob\u002Fmain\u002Fdocs\u002Finstall\u002Fone-click-install-release-copy.en.md) |\n| **同一入口内的公开版本** | `完整版 + 可定制治理` \u002F `框架版 + 可定制治理` |\n| **结果** | 在一处选择主机、操作和版本，然后复制匹配的提示 |\n\n\u003C\u002Fdiv>\n\n### 自定义：添加您自己的技能\n\n→ [自定义工作流与技能导入指南](https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\u002Fblob\u002Fmain\u002Fdocs\u002Finstall\u002Fcustom-workflow-onboarding.en.md)\n\n## 📦 站在巨人的肩膀上\n\n_这些能力并非从零开始构建。VibeSkills 的基础是将最佳开源解决方案持续集成到一个受治理的系统中。_\n\n我们深知，孤立地开发无法跟上快速发展的 AI 领域的步伐。VibeSkills 的核心优势在于不断吸收开源社区中最成熟的方法和架构，并将其纳入统一的治理与编排体系之下。\n\n> 🙏 **特别感谢与致谢**\n>\n> 本项目持续集成、吸收并治理以下优秀开源项目的精华：\n>\n> `superpower` · `claude-scientific-skills` · `get-shit-done` · `aios-core` · `OpenSpec` · `ralph-claude-code` · `SuperClaude_Framework` · `spec-kit` · `Agent-S` · `mem0` · `scrapling` · `claude-flow` · `serena` · `everything-claude-code` · `DeepAgent` 等\n>\n> _感谢各位作者的慷慨贡献——没有这些杰出的星光，VibeSkills 就不会存在。我们已尽力对所有被吸收的仓库进行恰当的署名和致谢。如有遗漏，请提交 Issue，我们将及时更正。_\n\n\u003Cbr\u002F>\n\n---\n\n\n## 🚀 开始使用\n\n_现在您已经了解了这是什么。接下来只需一条提示：_\n\n> ⚠️ **调用说明**：本项目采用 **技能格式架构**。请通过您的主机环境的技能调用方式来调用它——**不要**将其作为独立的 CLI 程序运行。\n\n\u003Cbr\u002F>\n\n\u003Cdiv align=\"center\">\n\n| 主机环境 | 调用方式 | 示例 |\n|:---:|:---:|:---|\n| **Claude Code** | `\u002Fvibe` | `我想让你设计一个 XXX \u002Fvibe` |\n| **Codex** | `$vibe` | `我想让你设计一个 XXXX $vibe` |\n| **OpenCode** | `\u002Fvibe` | `使用 vibe 技能来规划这次变更。` |\n| **OpenClaw** | 技能入口 | 参阅主机文档 |\n| **Cursor \u002F Windsurf** | 技能入口 | 参阅各平台的技能文档 |\n\n\u003C\u002Fdiv>\n\n\u003Cbr\u002F>\n\n> 💡 **提示**：为使每条消息都处于 VibeSkills 的受治理工作流中，请在每条消息后附加 `$vibe` 或 `\u002Fvibe`。未使用调用语法的消息将被视为受治理运行时之外的普通请求。\n\n**当前支持的平台**：`codex`（最完整的受治理路径）· `claude-code` · `cursor` · `windsurf`（支持安装使用路径，并集成了运行时适配器）· `openclaw`（预览运行时核心路径）· `opencode`（预览适配器路径）\n\n\u003Cbr\u002F>\n\n---\n\n\u003Cdetails>\n\u003Csummary>\u003Cb>📚 文档与安装指南（点击展开）\u003C\u002Fb>\u003C\u002Fsummary>\n\n\u003Cbr\u002F>\n\n**理解系统**\n\n- 📖 [系统架构与哲学](https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\u002Fblob\u002Fmain\u002Fdocs\u002Fquick-start.en.md)\n- 📜 [VibeSkills 宣言](https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\u002Fblob\u002Fmain\u002Fdocs\u002Fmanifesto.en.md)\n\n**安装与配置**\n\n- ⚡️ [基于提示的安装（推荐）](https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\u002Fblob\u002Fmain\u002Fdocs\u002Finstall\u002Fone-click-install-release-copy.en.md)\n- 🧩 [自定义工作流导入](https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\u002Fblob\u002Fmain\u002Fdocs\u002Finstall\u002Fcustom-workflow-onboarding.en.md)\n- 📄 [OpenClaw 主机注意事项](https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\u002Fblob\u002Fmain\u002Fdocs\u002Finstall\u002Fopenclaw-path.en.md)\n- 📄 [OpenCode 主机注意事项](https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\u002Fblob\u002Fmain\u002Fdocs\u002Finstall\u002Fopencode-path.en.md)\n- 📁 [手动复制安装（离线）](https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\u002Fblob\u002Fmain\u002Fdocs\u002Finstall\u002Fmanual-copy-install.en.md)\n- 🛠 [高级安装命令参考](https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\u002Fblob\u002Fmain\u002Fdocs\u002Finstall\u002Frecommended-full-path.en.md)\n- 🧊 [冷启动及其他环境](https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\u002Fblob\u002Fmain\u002Fdocs\u002Fcold-start-install-paths.en.md)\n\n\u003C\u002Fdetails>\n\n\u003Cbr\u002F>\n\n\u003Cdiv align=\"center\">\n\n### 🤝 加入社区 · 共同建设\n\n试试看吧！如果您有任何问题、想法或建议，欢迎随时提交 Issue——我会认真对待每一条反馈并不断改进。\n\n\u003Cbr\u002F>\n\n**本项目完全开源，欢迎一切贡献！**\n\n无论是修复 bug、提升性能、增加功能，还是改进文档，每一个 PR 都将受到衷心的感谢。\n\n```\nFork → Modify → Pull Request → Merge ✅\n```\n\n\u003Cbr\u002F>\n\n> ⭐ 如果本项目对您有所帮助，一颗 **Star** 就是对我们最大的支持！\n> 您的支持就是驱动这头核动力驴子的浓缩铀 🫏\n\n\u003Cbr\u002F>\n\n感谢 **LinuxDo** 社区的支持！\n\n[![LinuxDo](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FCommunity-LinuxDo-blue?style=for-the-badge)](https:\u002F\u002Flinux.do\u002F)\n\n技术讨论、AI 前沿、AI 经验分享——尽在 Linuxdo！\n\n\u003C\u002Fdiv>\n\n\u003Cbr\u002F>\n\n---\n\n## 星标历史\n\u003Cdiv align=\"center\">\n\u003Ca href=\"https:\u002F\u002Fwww.star-history.com\u002F?repos=foryourhealth111-pixel%2FVibe-Skills&type=date&legend=top-left\">\n \u003Cpicture>\n   \u003Csource media=\"(prefers-color-scheme: dark)\" srcset=\"https:\u002F\u002Fapi.star-history.com\u002Fimage?repos=foryourhealth111-pixel\u002FVibe-Skills&type=date&theme=dark&legend=top-left\" \u002F>\n   \u003Csource media=\"(prefers-color-scheme: light)\" srcset=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fforyourhealth111-pixel_Vibe-Skills_readme_c7193a8d707c.png\" \u002F>\n   \u003Cimg alt=\"星标历史图表\" src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fforyourhealth111-pixel_Vibe-Skills_readme_c7193a8d707c.png\" \u002F>\n \u003C\u002Fpicture>\n\u003C\u002Fa>\n\n---\n\n\u003Cdiv align=\"center\">\n  \u003Cp>\u003Ci>将实际工作中最容易失控的部分转化为一个更易调用、更易管理且长期可维护的系统。\u003C\u002Fi>\u003C\u002Fp>\n  \u003Cbr\u002F>\n  \u003Csub>用心打造  ·  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\">GitHub\u003C\u002Fa>  ·  \u003Ca href=\".\u002FREADME.zh.md\">中文\u003C\u002Fa>\u003C\u002Fsub>\n\u003C\u002Fdiv>","# Vibe-Skills 快速上手指南\n\nVibe-Skills 不仅仅是一个技能集合，更是你的**个人 AI 操作系统**。它通过智能路由和治理机制，协调 340+ 个技能（涵盖编码、研究、数据科学与创意工作），确保 AI 任务高效、可靠地交付。\n\n## 🛠️ 环境准备\n\n在开始之前，请确保你的开发环境满足以下要求：\n\n*   **操作系统**：Linux, macOS 或 Windows (WSL2 推荐)。\n*   **前置依赖**：\n    *   **Node.js**: 建议版本 v18 或更高。\n    *   **Git**: 用于克隆仓库和管理版本。\n    *   **AI 编辑器\u002F环境**: 支持 Skills 协议的工具，如 **Cursor**, **Windsurf**, **Claude Code**, **OpenCode** 等。\n    *   **MCP 兼容性**: 原生支持 Model Context Protocol (MCP)。\n\n> 💡 **提示**：本工具旨在降低高级 AI 编程（Vibecoding）的使用门槛，无论是否有深厚的编程背景，均可直接使用。\n\n## 📥 安装步骤\n\n### 1. 克隆仓库\n打开终端，运行以下命令将 Vibe-Skills 克隆到本地：\n\n```bash\ngit clone https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills.git\ncd Vibe-Skills\n```\n\n> 🚀 **国内加速方案**：如果访问 GitHub 较慢，可使用国内镜像源加速克隆：\n> ```bash\n> git clone https:\u002F\u002Fgitee.com\u002Fmirror\u002FVibe-Skills.git  # 示例镜像地址，请替换为实际可用的国内镜像\n> # 或者使用 GitCode 镜像\n> git clone https:\u002F\u002Fgitcode.com\u002Fforyourhealth111-pixel\u002FVibe-Skills.git\n> cd Vibe-Skills\n> ```\n\n### 2. 一键安装与配置\n执行官方提供的一键安装脚本，自动完成运行时环境（VCO Runtime）的配置和技能注册：\n\n```bash\n.\u002Finstall.sh\n```\n\n*Windows 用户请使用:*\n```powershell\n.\\install.ps1\n```\n\n安装完成后，系统会自动初始化 **340+** 个技能模块，并配置好语义化目录结构。\n\n## 🚀 基本使用\n\n安装完成后，你无需记忆复杂的技能列表，只需通过统一的指令入口即可触发智能路由。\n\n### 核心指令\n在你的 AI 编辑器（如 Cursor\u002FWindsurf）的聊天窗口或终端中，输入以下任一指令：\n\n```text\n\u002Fvibe\n```\n或\n```text\n$vibe\n```\n\n### 使用示例\n\n**场景**：你需要为一个现有项目添加用户登录功能，并进行测试。\n\n1.  **发起请求**：\n    在 AI 对话框中输入：\n    ```text\n    \u002Fvibe 请为当前项目添加基于 JWT 的用户登录功能，包含注册、登录接口及相应的单元测试。\n    ```\n\n2.  **自动流程执行**：\n    Vibe-Skills 将自动执行以下受治理的工作流：\n    *   **🧠 智能路由**：识别意图，匹配 `auth-implementation` 和 `tdd-guide` 等技能。\n    *   **📝 需求澄清**：自动确认技术栈和验收标准（若信息不足会主动询问）。\n    *   **🏗️ 架构规划**：生成分步实施计划。\n    *   **⚙️ 分级执行**：根据任务复杂度自动选择 **M** (简单)、**L** (中等) 或 **XL** (多代理并行) 模式执行。\n    *   **✅ 验证交付**：运行测试并生成证明包（Proof Bundle），确保代码无误。\n\n3.  **指定执行等级（可选）**：\n    对于大型重构任务，可显式指定使用 XL 级多代理工作流：\n    ```text\n    \u002Fvibe 重构整个数据层模块，启动 XL 级工作流\n    ```\n\n### 工作流程概览\n```text\n安装 → 输入 \u002Fvibe 或 $vibe → 智能路由 → M\u002FL\u002FXL 分级执行 → 治理验证 → ✅ 交付\n```\n\n现在，你可以开始体验由 340+ 技能协同驱动的可靠 AI 开发工作流了！","某全栈开发者需要在两天内为一个初创项目搭建包含后端 API、数据库架构及基础前端页面的最小可行性产品（MVP），同时确保代码符合测试驱动开发（TDD）规范。\n\n### 没有 Vibe-Skills 时\n- **上下文频繁切换**：开发者需在不同 AI 对话窗口中分别请求写代码、查文档和生成测试用例，导致思路中断，难以维持统一的项目语境。\n- **流程缺乏管控**：AI 直接生成代码而跳过规划与验证步骤，产出的代码往往缺少单元测试或不符合既定架构规范，后期返工成本高。\n- **技能调用碎片化**：手动拼接“代码审查”、“数据库设计”等独立指令，难以让多个专业能力（如 TDD 指南与 API 生成）协同工作，效率低下。\n- **记忆断层严重**：随着对话深入，AI 逐渐遗忘早期的技术选型约束，导致新生成的代码模块与原有逻辑冲突。\n\n### 使用 Vibe-Skills 后\n- **一键智能路由**：只需输入 `\u002Fvibe` 并描述需求，Vibe-Skills 的 Canonical Router 自动识别任务，按需调用编码、研究与测试等 340+ 技能模块，无需人工干预。\n- **治理化执行流程**：系统强制遵循“澄清→规划→执行→验证”的标准流程，确保在编写 API 前自动生成 TDD 测试用例，从源头保障代码质量。\n- **多技能无缝协同**：Vibe-Skills 将数据库设计与前端生成技能串联，保持上下文一致，一次性交付前后端贯通且经过验证的完整功能模块。\n- **持久化记忆系统**：内置记忆机制全程记录技术决策与架构约束，即使在进行复杂的多步骤任务时，也能确保所有输出严格符合初始设定。\n\nVibe-Skills 通过将分散的 AI 能力整合为受控的操作系统级工作流，让开发者从繁琐的指令拼凑中解放，专注于核心业务逻辑的创新。","https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fforyourhealth111-pixel_Vibe-Skills_947a6dc2.png","foryourhealth111-pixel","QingFeng Li","https:\u002F\u002Foss.gittoolsai.com\u002Favatars\u002Fforyourhealth111-pixel_5b99b190.png","Romanticism, the perfectionist geek",null,"foryourhealth111@gmail.com","https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel",[83,87,91,95,99,103,107,111,115,118],{"name":84,"color":85,"percentage":86},"Python","#3572A5",52.8,{"name":88,"color":89,"percentage":90},"PowerShell","#012456",36.8,{"name":92,"color":93,"percentage":94},"TeX","#3D6117",6.4,{"name":96,"color":97,"percentage":98},"Shell","#89e051",2.4,{"name":100,"color":101,"percentage":102},"JavaScript","#f1e05a",0.8,{"name":104,"color":105,"percentage":106},"HTML","#e34c26",0.4,{"name":108,"color":109,"percentage":110},"Jupyter Notebook","#DA5B0B",0.2,{"name":112,"color":113,"percentage":114},"TypeScript","#3178c6",0.1,{"name":116,"color":117,"percentage":114},"Swift","#F05138",{"name":119,"color":120,"percentage":121},"CSS","#663399",0,972,76,"2026-04-02T20:51:36","Apache-2.0","未说明",{"notes":128,"python":126,"dependencies":129},"该工具并非传统的本地运行 AI 模型，而是一个技能编排系统（Vibe Code Orchestrator），旨在管理 340+ 个技能模块。它不依赖特定的 GPU、显存或 Python 版本在本地运行核心逻辑，而是作为中间层连接用户与外部支持的 AI 环境（如 Cursor, Windsurf, Claude Code 等）。只要您的 AI 编辑器支持 Skills 协议或 MCP 协议即可使用。安装后通过 `\u002Fvibe` 或 `$vibe` 命令触发智能路由和工作流治理。",[130,131,132,133,134,135,136],"Claude Code","Codex","Windsurf","OpenClaw","OpenCode","Cursor","MCP (Model Context Protocol)",[13,26,15],[139,140,141,142,143,144,145,146,147,148,149,150,151,152,153,154,155,156,157,158],"ai-agents","ai-workflow","automation","codex","developer-tools","mcp","skills","ai-skills","claude-code","llm","vibe-coding","windsurf","agent-skills","agentic-coding","anthropic","claude","claude-skills","cursor","opencode","vibecoding","2026-03-27T02:49:30.150509","2026-04-06T10:26:25.997866",[162,167,172,177,182,187],{"id":163,"question_zh":164,"answer_zh":165,"source_url":166},11025,"如何为 Cursor 编辑器添加代理适配支持？","项目已通过 PR #13 添加了 `cursor` 预览适配器支持，并将其集成到安装、检查和引导流程中。该方案已针对临时目标和实际的 Cursor 目标路径进行了本地验证。用户可以直接使用更新后的版本来启用对 Cursor 的支持。","https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\u002Fissues\u002F6",{"id":168,"question_zh":169,"answer_zh":170,"source_url":171},11026,"会话产生的过程产物（如文档）默认存储在哪里？如何区分不同工作空间的文件？","目前会话的过程产物存储在 `~\u002F.codex` 目录下。为了更好地支持多项目开发和工作空间隔离，建议（并正在实施）将 `~\u002F.codex` 仅用于存放 skill 安装内容，而实际使用过程中产生的文件（如 doc 等）将存储在工作空间根目录下的 `.vibe` 文件夹中。这样不同工作空间默认即可实现文件隔离，类似于 Codex 桌面版和 Trae 的 spec 模式。","https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\u002Fissues\u002F88",{"id":173,"question_zh":174,"answer_zh":175,"source_url":176},11027,"在 Claude Code 中使用 \u002Fvibe 命令调用失效怎么办？","该问题是由于路由触发功能和 vibe 调用接口之间出现断裂导致的。维护者已定位并修复了此连接问题，现在 \u002Fvibe 命令应能正常调用。","https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\u002Fissues\u002F43",{"id":178,"question_zh":179,"answer_zh":180,"source_url":181},11028,"Codex 中出现重复的 Vibe Skill 条目如何处理？","这是一个已知问题，表现为 Codex 中显示了两个 Vibe Skill。维护者已经修复了该重复显示的问题，更新后应只显示一个正确的 Skill 条目。","https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\u002Fissues\u002F42",{"id":183,"question_zh":184,"answer_zh":185,"source_url":186},11029,"如何测试 AI 路由功能的连通性以确认配置是否正确？","针对路由层的 AI 意图分析功能，项目已增加了连通性测试途径。用户可以通过该测试功能验证配置是否生效，从而确认路由功能是否正常运作。","https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\u002Fissues\u002F33",{"id":188,"question_zh":189,"answer_zh":190,"source_url":191},11030,"项目支持哪些代码编辑器或 IDE 的集成？","项目已逐步添加了对多种编辑器的支持，包括：Cursor（通过预览适配器）、Windsurf、OpenCode、OpenClaw（龙虾）以及纯框架模式（无 Skills）。此外，还修复了 Claude Code 中的调用问题。用户可以查阅相关 Issue 或文档获取具体编辑器的安装和配置指南。","https:\u002F\u002Fgithub.com\u002Fforyourhealth111-pixel\u002FVibe-Skills\u002Fissues\u002F10",[193,198,203,208,213,218],{"id":194,"version":195,"summary_zh":196,"released_at":197},54198,"v2.3.55","# VCO 发布 v2.3.55\n\n- 日期：2026-03-30\n- 基础提交：f3ab6e5\n- 上一版本：`v2.3.54`\n\n## 亮点\n\n- 将统一的“仅限所有者”卸载界面正式纳入稳定版发布线。`uninstall.sh` 和 `uninstall.ps1` 现在通过相同的适配器驱动契约路由支持的主机，默认采用直接卸载方式，并且仅移除那些可通过安装账本、主机关闭机制或保守的遗留规则证明属于 Vibe 的内容。\n- 重新调整了主机安装、检查及运行时框架，围绕明确的“仅技能”和“侧车优先”的激活模式展开。支持的主机将通过原生技能调用保持 Vibe 可用，而无需在正常路径中由安装程序接管主机原生设置，从而降低跨主机破坏的风险。\n- 修复了 OpenCode 启动回归问题，方法是保留与现有 OpenCode 配置界面的兼容性，并缩小受管理的写入范围。现在，受管理的安装\u002F更新流程不会再将原本有效的 OpenCode 配置变为无法启动的状态。\n- 将内置意图建议与可选的向量差异嵌入拆分为独立的配置平面：`VCO_INTENT_ADVICE_*` 现在负责治理建议路径，而 `VCO_VECTOR_DIFF_*` 仍作为可选的嵌入通道，在缺失时会优雅降级。\n- 加强了 macOS 引导兼容性。活跃的 Shell 入口点不再依赖仅 Bash 4 才支持的 `mapfile`，并且 `install.sh`、`check.sh` 以及 `scripts\u002Fbootstrap\u002Fone-shot-setup.sh` 现在会在早期就以清晰的 Python 3.10+ 必备条件提示失败，而不是抛出具有误导性的下游错误。\n\n## 相较于 v2.3.54 的变化\n\n- `v2.3.54` 侧重于发布运维闭环、发布与安装状态的一致性，以及卸载治理界面的暴露。\n- `v2.3.55` 则将稳定版界面推进到下一组已合并的主机安全性修复上：包括仅技能的主机对齐、OpenCode 启动恢复、显式的 AI 配置键分离，以及 macOS Shell 和引导兼容性改进。\n- 主要区别并不在于新增的产品功能线，而在于更小的影响范围和更清晰的契约：安装与卸载现在对主机原生状态的干预更少，OpenCode 依然能够正常启动，AI 配置更加明确，而 macOS 必备条件未满足的情况也会在实际发生的位置被准确诊断出来。\n\n## 验证说明\n\n- `pytest -q tests\u002Fruntime_neutral\u002Ftest_uninstall_vgo_adapter.py tests\u002Fruntime_neutral\u002Ftest_installed_runtime_uninstall.py tests\u002Fruntime_neutral\u002Ftest_installed_runtime_scripts.py -k \"install_ledger or test_\"` -> `31 个测试通过，10 个子测试通过`\n- `pytest -q tests\u002Fruntime_neutral\u002Ftest_opencode_managed_preview.py tests\u002Fruntime_neutral\u002Ftest_bootstrap_doctor.py tests\u002Fruntime_neutral\u002Ftest_router_ai_connectivity_probe.py tests\u002Fruntime_neutral\u002Ftest_shell_entrypoint_compatibility.py` -> `18 个测试通过`\n- `bash -n install.sh` -> `通过`\n- `bash -n check.sh` -> `通过`\n- `bash -n scripts\u002Fbootstrap\u002Fone-shot-setup.sh` -> `通过`\n- `pwsh -NoProfile -File scripts\u002Fverify\u002Fvibe-version-consistency-gate.ps1` -> `通过`\n- `pwsh -NoProfile -File scri","2026-03-30T16:03:25",{"id":199,"version":200,"summary_zh":201,"released_at":202},54199,"v2.3.54","# VCO 发布 v2.3.54\n\n- 日期：2026-03-30\n- 基础提交：1a049f1\n- 上一版本：`v2.3.53`\n\n## 亮点\n\n- 弥补了 `v2.3.53` 留下的发布表面与实际状态之间的差距：`release-cut.ps1` 现在成为版本治理、变更日志\u002F账本写入、`docs\u002Freleases\u002FREADME.md`、分发清单 `source_release` 更新，以及发布应用过程中捆绑包\u002F嵌套捆绑包同步的权威路径。\n- 将运行时契约债务相关工作提升为有文档记录的基线，而非分散的隐式结构。运行时包投影、模式引用、契约黄金标准和主机\u002F运行时投影测试，如今为发布提供了稳定的证明表面，便于未来的重构。\n- 完成了当前目标 Fixture 家族的跟踪输出边界迁移，并确保兼容性无误：规范基线已迁移到受控的 Fixture\u002F引用表面，而发布验证持续证明未引入回归。\n- 将安装时生成的嵌套兼容性正式纳入支持的发布\u002F安装路径，而非依赖人工维护的假设；同时确保已安装运行时的行为无需依赖仓库级别的适配器注册表即可正常工作。\n- 加强了发布事实本身：发布说明现设有专门的质量门控，`v2.3.54` 替换了过时的 `v2.3.53` 当前发布界面，且发布的摘要内容与该工作树中实际交付的代码完全一致。\n\n## 相较于 v2.3.53 的变化\n\n- `v2.3.53` 主要描述了专用调度闭包、Windows PowerShell 主机解析以及清理性措辞。\n- `v2.3.54` 才真正包含了该次切割之后完成的后续非回归修复改进：发布操作员闭环、运行时契约模式基线化、输出边界迁移、安装时生成的嵌套兼容性，以及发布说明事实的强化。\n- 这次发布本质上是对发布表面的修正，而非单纯的代码发布。主要区别不在于新增面向用户的特性，而在于发布\u002FREADME\u002F版本\u002F清单及运行时证明现在都指向同一个经过修复的仓库状态。\n\n## 验证说明\n\n- `pytest -q tests\u002Fruntime_neutral\u002Ftest_release_cut_operator.py` -> `2 passed`\n- `pytest -q tests\u002Fruntime_neutral\u002Ftest_release_notes_quality.py` -> `4 passed`\n- `pytest -q tests\u002Fruntime_neutral\u002Ftest_outputs_boundary_migration.py` -> `2 passed`\n- `pytest -q tests\u002Fruntime_neutral\u002Ftest_install_generated_nested_bundled.py tests\u002Fruntime_neutral\u002Ftest_generated_nested_bundled.py` -> `5 passed`\n- `pytest -q tests\u002Fruntime_neutral\u002Ftest_runtime_contract_schema.py tests\u002Fruntime_neutral\u002Ftest_runtime_contract_goldens.py` -> `13 passed`\n- `pytest -q tests\u002Fruntime_neutral\u002Ftest_installed_runtime_scripts.py -k \"nested_runtime_skill_entrypoints_sanitized or installed_shell_scripts_work_without_repo_level_adapter_registry\"` -> `1 passed, 11 deselected`\n- `pwsh -NoProfile -File scripts\u002Fverify\u002Fvibe-version-consistency-gate.ps1` -> `PASS`\n- `pwsh -NoProfile -File scripts\u002Fverify\u002Fvibe-dist-manifest-gate.ps1` -> `PASS`\n- ","2026-03-30T05:23:31",{"id":204,"version":205,"summary_zh":206,"released_at":207},54200,"v2.3.53","# VCO 发布 v2.3.53\n\n- 日期：2026-03-30\n- 基础提交：4f5676f\n- 上一版本：`v2.3.52`\n\n## 亮点\n\n- 通过明确的自定义准入处理，弥合了受管专家调度中的差距，并恢复了缺失的委派通道\u002F主机适配器元数据交接。路由器准入、确认 UI、运行时输入包、委派通道负载以及专家执行现在都承载相同的有界调度事实，而非依赖于较弱的隐含行为。\n- 在安装、检查和引导界面中强化了 Windows PowerShell 主机解析，并在 Claude Code、Cursor、Windsurf、OpenClaw 和 OpenCode 等适配器通道上收紧了受管主机安装的保证。\n- 优化了清理阶段相关措辞和策略，使公开文档及运行时政策不再过度强调清理阶段所能保证的内容，尤其是在“绝对无残留”这类声明方面。\n\n## 相较于 v2.3.52 的变化\n\n- `v2.3.52` 将受管内存激活纳入标准的六阶段运行时流程，并使后端内存活动可被观测。\n- `v2.3.53` 则将重点重新聚焦于面向发布的可靠性和事实一致性：专家准入、专家调度闭环、主机安装保证以及清理安全性声明，如今在路由器、运行时、安装和适配器等各个界面上更加紧密地对齐。\n- 重要的区别并不在于新增产品线，而在于受管执行和面向主机的入口点现在做出的承诺更少且可验证，同时对于已作出的承诺则提供了更强的显式证明。\n\n## 验证说明\n\n- 运行时\u002F专家调度覆盖率：\n  - `pytest -q tests\u002Fruntime_neutral\u002Ftest_custom_admission_bridge.py tests\u002Fruntime_neutral\u002Ftest_multi_host_specialist_execution.py`\n- 安装\u002F主机解析覆盖率：\n  - `pytest -q tests\u002Fruntime_neutral\u002Ftest_installed_runtime_scripts.py`\n- 发布界面与版本一致性：\n  - `git diff --check`\n  - `pwsh -NoProfile -File scripts\u002Fverify\u002Fvibe-version-consistency-gate.ps1`\n  - `pwsh -NoProfile -File scripts\u002Fverify\u002Fvibe-dist-manifest-gate.ps1`\n  - `pwsh -NoProfile -File scripts\u002Fverify\u002Fvibe-specialist-dispatch-closure-gate.ps1`\n\n## 迁移说明\n\n- 在标准 Windows PowerShell 主机上运行的用户，在常规支持的配置下无需进行临时性的 Shell 覆盖，即可获得更可靠的安装和运行时检测。\n- 具备专家能力的受管运行现在会表现出更为严格的准入和调度行为。请将这些记录和检查结果视为权威依据，而不要假定旧有的隐式调度行为仍然适用。\n- 本次发布提升了调度闭环和事实一致性的水平，但并未宣称超出当前适配器和发行版文档所描述范围之外的任何新的广泛主机原生支持层级。","2026-03-29T16:47:43",{"id":209,"version":210,"summary_zh":211,"released_at":212},54201,"v2.3.52","# VCO 发布 v2.3.52\n\n- 日期：2026-03-29\n- 基础提交：870fd20\n- 上一版本：`v2.3.51`\n\n## 亮点\n\n- 在正常的六阶段 `vibe` 管治运行时中，实现了阶段感知的内存激活机制。管治运行现在会生成 `memory-activation-report.json` 和 `memory-activation-report.md` 文件，将有界的内存上下文注入到需求\u002F计划工件中，并在各阶段的收据中保留激活路径的可见性。\n- 为 `Serena`、`ruflo` 和 `Cognee` 添加了真实的管治后端适配器调用。启用后，运行时会将请求\u002F响应工件写入会话的 `memory-backend\u002F` 目录下，而不再仅记录建议性的所有权策略。\n- 保持内存契约的狭窄和明确，而非将其扩展为不受控制的后台内存层：\n  - `Serena` 仍用于明确的项目决策；\n  - `Cognee` 仍用于有界的关联召回与摄取；\n  - `ruflo` 仍以 XL 范围限定，用于 `plan_execute` 中的里程碑和交接连续性；\n  - `knowledge-steward` 仍然是一项显式技能，而非由运行时自动捕获；\n  - 清理折叠是本地管治折叠工件，而非对 `deepagent-memory-fold` 技能的静默自动调用。\n\n## 与 v2.3.51 的变化\n\n- `v2.3.51` 将交付验收纳入主管治运行链，并强化了专家治理的封闭性。\n- `v2.3.52` 则将内存层面从主要的策略\u002F路由建议，推进到具有真实后端读写路径和运行时工件的可观测阶段绑定运行行为。\n- 重要的区别不在于“到处都有更多内存”，而在于管治运行现在能够展示内存被读取的位置、写入的位置，以及运行时为何有意拒绝扩大内存范围。\n\n## 验证说明\n\n- 内存激活\u002F运行时后端覆盖：\n  - `pytest -q tests\u002Fruntime_neutral\u002Ftest_memory_runtime_activation.py`\n- 管治运行时\u002F拓扑\u002F层级回归覆盖：\n  - `pytest -q tests\u002Fruntime_neutral\u002Ftest_governed_runtime_bridge.py tests\u002Fruntime_neutral\u002Ftest_l_xl_native_execution_topology.py tests\u002Fruntime_neutral\u002Ftest_root_child_hierarchy_bridge.py`\n- 发布表面整洁度检查：\n  - `git diff --check`\n  - `pwsh -NoProfile -File scripts\u002Fverify\u002Fvibe-version-consistency-gate.ps1`\n  - `pwsh -NoProfile -File scripts\u002Fverify\u002Fvibe-dist-manifest-gate.ps1`\n  - `pwsh -NoProfile -File scripts\u002Fverify\u002Fvibe-nested-bundled-parity-gate.ps1`\n\n## 迁移说明\n\n- 运维人员应预期在 `outputs\u002Fruntime\u002Fvibe-sessions\u002F\u003Crun-id>\u002F` 下出现新的运行时工件，尤其是内存激活报告系列；当启用后端通道时，还会在 `memory-backend\u002F` 目录下生成请求\u002F响应收据。\n- 本版本证明了管治内存激活及有界后端集成的能力，但并未宣称所有与内存相关的技能都会在后台静默自动运行。\n- `knowledge-steward` 仍需用户明确意图，而运行时清理折叠应被视为本地管治上下文的压缩，而非作为一项……","2026-03-29T02:24:44",{"id":214,"version":215,"summary_zh":216,"released_at":217},54202,"v2.3.51","# VCO 发布 v2.3.51\n\n- 日期：2026-03-28\n- 基础提交：18e6b9c\n- 上一版本：`v2.3.50`\n\n## 亮点\n\n- 将下游交付验收从外部验证层移至由 `vibe` 治理的正常运行时主链中。治理运行现在会直接在主要需求界面上冻结产品验收标准、人工抽查、完成语言政策以及交付真相合约。\n- 扩展了治理计划与收尾路径，使 `xl_plan` 记录交付验收计划，而 `phase_cleanup` 则为每次运行生成一份 `delivery-acceptance-report.json` 文件。这使得完成语言降级成为实际运行路径的一部分，而非事后审查惯例。\n- 完成了近期在 `main` 分支上实施的专家治理序列：阶段限定的专家派遣、根节点批准下的子通道同轮自动吸收，以及更强的原生专家失败防护机制。这提升了 `vibe` 在不放弃运行时权威的前提下使用专家协助的能力。\n- 修复了 Windows 环境下的专家运行时交接问题，确保原生专家执行在当前 Windows 环境中仍可正常运行。\n\n## 相较于 v2.3.50 的变化\n\n- `v2.3.50` 主要关注安装真相、路由器 AI 连通性验证，以及主机适配器和安装界面的收尾工作。\n- `v2.3.51` 则将重心转移到治理执行路径本身。最大的不同在于，代码库不再仅仅强调“交付真相很重要”；如今，正常的 `vibe` 运行时会将其自身产物中记录并承载这一真相。\n- 旧版本着重提升就绪状态披露与验证入口；而新版本则提高了在治理运行结束后声明工作完成时的诚实度门槛。\n\n## 验证说明\n\n- 主链交付验收覆盖率：\n  - `pytest -q tests\u002Fruntime_neutral\u002Ftest_runtime_delivery_acceptance.py tests\u002Fruntime_neutral\u002Ftest_workflow_acceptance_runner.py tests\u002Fruntime_neutral\u002Ftest_release_truth_gate.py`\n- 治理运行时 \u002F 层次结构 \u002F 拓扑覆盖：\n  - `pytest -q tests\u002Fruntime_neutral\u002Ftest_governed_runtime_bridge.py tests\u002Fruntime_neutral\u002Ftest_l_xl_native_execution_topology.py tests\u002Fruntime_neutral\u002Ftest_root_child_hierarchy_bridge.py`\n- 发布表面整洁性检查：\n  - `git diff --check`\n\n## 迁移说明\n\n- 运维人员应预期在正常治理运行后，完成语句的要求将更加严格。干净的运行时\u002F流程收尾不再被视为等同于下游项目已成功交付。\n- 正常 `vibe` 会话产物现包含一个 `delivery-acceptance-report.json` 文件，以及位于 `outputs\u002Fruntime\u002Fvibe-sessions\u002F\u003Crun-id>\u002F` 下的配套 Markdown 摘要。\n- 本次发布提升了真相性和收尾语义的准确性。但并未宣称所有主机 UI 现在都强制实施绝对不可跳过的最终文本关口；其权威性变化在于，治理运行时本身现在会记录并承载交付真相的结果。","2026-03-28T14:47:23",{"id":219,"version":220,"summary_zh":221,"released_at":222},54203,"v2.3.50","# VCO 版本 v2.3.50\n\n- 日期：2026-03-26\n- 基础提交：d5da111\n\n## 亮点\n\n- 为治理建议路径新增了路由器 AI 连通性探测器，包括 PowerShell 入口、运行时无关的 Python 探测脚本、结构化状态工件，以及用于区分本地安装完成与在线治理就绪状态的快速检查机制。\n- 加强了 LLM 加速叠加层对可选字段的处理逻辑，确保在提供商侧字段缺失或部分填充时，验证和回退行为仍能保持稳定。\n- 扩展了当前主分支中真实主机适配器的支持范围：规范化并完善了 OpenClaw 运行时核心流程，新增了 OpenCode 预览适配器及安装界面，并使 Cursor\u002FWindsurf 预览支持与最新的安装\u002F支持矩阵保持一致。\n- 将公开安装入口整合为单一入口，提升了安装文档的双语一致性，并将默认的 Windows 验证示例改为 `powershell.exe`，以避免原生 Windows 环境因依赖 `pwsh` 而受阻。\n\n## 验证说明\n\n- 版本治理闭环：\n  - `pwsh -NoProfile -ExecutionPolicy Bypass -File .\u002Fscripts\u002Fverify\u002Fvibe-version-consistency-gate.ps1`\n  - `pwsh -NoProfile -ExecutionPolicy Bypass -File .\u002Fscripts\u002Fverify\u002Fvibe-version-packaging-gate.ps1`\n- 路由器 AI 连通性探测器覆盖：\n  - `python3 -m py_compile scripts\u002Fverify\u002Fruntime_neutral\u002Frouter_ai_connectivity_probe.py`\n  - `python3 -m pytest -q tests\u002Fruntime_neutral\u002Ftest_bootstrap_doctor.py tests\u002Fruntime_neutral\u002Ftest_router_ai_connectivity_probe.py`\n  - `pwsh -NoProfile -ExecutionPolicy Bypass -File .\u002Fscripts\u002Fverify\u002Fvibe-llm-acceleration-overlay-gate.ps1`\n- 发布表面整洁度检查：\n  - `git diff --check`\n\n## 迁移说明\n\n- Windows 管理员应使用 `powershell.exe -NoProfile -ExecutionPolicy Bypass -File ...` 作为默认的验证入口，除非已安装 PowerShell 7 (`pwsh`)。\n- 当前的公开安装指南已集中于 `docs\u002Finstall\u002Fone-click-install-release-copy*.md` 下的单一入口文档；旧有的分散式安装入口不应再被视为主要的入门界面。\n- OpenClaw、OpenCode、Cursor 和 Windsurf 仍按照当前支持矩阵归属为真实预览\u002F运行时核心治理通道。本次发布旨在提升安装的真实性、验证的清晰度以及适配器的完整性，但并未宣称实现全面的主机原生完全兼容。","2026-03-26T12:56:14"]