vibesop 5.3.1

AI SkillOS — discover skills, route queries, orchestrate tasks, manage lifecycles. Unified dashboard, interactive cleanup, community sharing.

VibeSOP

AI 辅助开发的技能操作系统

The Skill Operating System for AI-assisted development

---

愿景 Vision

不再记忆命令，只需表达意图。 不再猜测工具，智能匹配最佳。 不再学习平台，一次掌握所有。

No more memorizing commands. Just express your intent. No more guessing tools. Intelligent matching finds the best. No more learning platforms. Master them all at once.

---

什么是 VibeSOP？ What is VibeSOP?

VibeSOP 是 SkillOS（技能操作系统）——管理技能的全生命周期：

VibeSOP is a Skill Operating System — managing the full lifecycle of skills:

技能全生命周期管理（SkillOS 定位）Skill Lifecycle Management

• 发现与安装：一键安装技能，自动安全审计，零配置 Discovery & Installation — one-click install, auto security audit, zero config

• 智能路由：理解意图，从 50+ 技能中匹配最佳 Intelligent Routing — understand intent, match the best from 50+ skills

• 任务编排：复杂请求自动分解，串行/并行执行计划 Task Orchestration — decompose complex requests, serial/parallel execution plans

• 生命周期管理：启禁用、作用域隔离、质量评估、自动淘汰 Lifecycle Management — enable/disable, scope isolation, quality evaluation, auto-deprecation

• 跨平台适配：一套技能定义，所有 AI Agent 通用 Cross-Platform — one skill definition, works with all AI Agents

VibeSOP 定位: VibeSOP 是 SkillOS。它管理技能的全生命周期：发现 → 安装 → 路由 → 编排 → 评估 → 保留/淘汰。 技能的实际执行由 AI Agent（Claude Code, Cursor, OpenCode）完成。

Note: VibeSOP is a Skill Operating System. It manages the full skill lifecycle: discovery → installation → routing → orchestration → evaluation → retention/deprecation. Actual skill execution is done by AI Agents (Claude Code, Cursor, OpenCode).

📖 Read our philosophy: docs/PHILOSOPHY.md | 中文版

---

核心价值 Core Values

Discovery over Execution (发现 > 执行)

找到正确的工具比执行更重要。AI 工具已经足够强大，真正的问题是：找到正确的工具。

Finding the right tool is more important than executing it. AI tools are already powerful enough. The real problem is: finding the right tool.

Orchestration over Single-Skill (编排 > 单技能)

真实世界的请求往往是复合的。VibeSOP 能够分解复杂意图，编排多个技能协同工作。

Real-world requests are composite. VibeSOP decomposes complex intents and orchestrates multiple skills working together.

Lifecycle over Accumulation (生命周期 > 堆积)

技能应该被管理，而不是无限堆积。启用/禁用、作用域隔离、质量评估、自动淘汰——让技能生态保持健康。

Skills should be managed, not infinitely accumulated. Enable/disable, scope isolation, quality evaluation, auto-deprecation — keeping the skill ecosystem healthy.

Matching over Guessing (匹配 > 猜测)

理解意图比记忆命令更重要。你记不住 50+ 个技能的命令，但你可以自然地表达你想做什么。

Understanding intent is more important than memorizing commands. You can't remember 50+ skill commands, but you can naturally express what you want to do.

Open over Closed (开放 > 封闭)

开放生态比封闭系统更有价值。VibeSOP 不绑定任何平台，你可以使用任何 AI 工具。

An open ecosystem is more valuable than a closed system. VibeSOP doesn't bind to any platform — you can use any AI tool.

---

⚡ 快速开始 Quick Start

🚀 一键安装技能（核心特性）

从 8 个手动步骤 → 1 条命令，98% 时间节省！

# 安装技能 - 自动配置，零学习曲线
vibe skill add tushare

# 系统自动完成：
# ✅ 检测技能类型
# ✅ 安全审计
# ✅ 智能配置路由规则
# ✅ 自动设置优先级
# ✅ 验证和同步

# 立即开始使用
vibe route "帮我获取茅台最近一年的股价"
# → AI 自动匹配到 tushare 技能（95% 置信度）

对比旧流程：

• ❌ 旧方式：30-60 分钟，8+ 手动步骤，40% 错误率
• ✅ 新方式：1-2 分钟，1 条命令，<5% 错误率

详见：智能技能安装文档

---

安装 VibeSOP

# Clone the repository
git clone https://github.com/nehcuh/vibesop-py.git
cd vibesop-py

# Install with uv (recommended - 10-100x faster than pip)
uv sync

# Or with pip
pip install -e .

第一次使用 First Use

# Single intent - routes to best skill
$ vibe route "帮我调试这个错误"

🔍 Routing Summary
─────────────────────────────
Selected     systematic-debugging
Confidence   95%
Layer        scenario
Duration     12.3ms

💡 Alternatives:
   • gstack/investigate (82%)
   • superpowers/debug (75%)

# Multi intent - automatically orchestrates
$ vibe route "分析架构并生成测试"

🔍 Routing Summary
─────────────────────────────
Mode         Orchestrated
Steps        2
Strategy     sequential

Plan:
  1. riper-workflow — 架构分析
  2. superpowers/test — 测试生成

[✅ Confirm] [✏️ Edit] [🔀 Single skill] [📝 Skip]

就这么简单！ VibeSOP 理解你的意图——无论是单一任务还是复杂多步骤请求。

That's it! VibeSOP understands your intent — whether it's a single task or a complex multi-step request.

⚠️ Important: VibeSOP requires its own LLM configuration

VibeSOP runs as a CLI subprocess and cannot reuse the host Agent's internal LLM (e.g., OpenCode or Claude Code's session model). You must configure a separate LLM API key or local Ollama service for VibeSOP. Without LLM, VibeSOP uses keyword/TF-IDF matching only, and long queries may fail to match any skill.

# Anthropic Claude (recommended)
export ANTHROPIC_API_KEY="sk-ant-..."
# or OpenAI
export OPENAI_API_KEY="sk-..."
# or local Ollama (zero cost, no data leaving your machine)
export VIBE_LLM_PROVIDER=ollama
export OLLAMA_BASE_URL=http://localhost:11434/v1
export OLLAMA_MODEL=qwen3:35b-a3b-mlx

---

为什么选择 VibeSOP？ Why VibeSOP?

问题 The Problem

AI 辅助开发工具正在爆发：

• Claude Code, Cursor, Continue.dev, Aider...
• 每个工具都有自己的命令和技能
• superpowers, gstack, omx 等技能包蓬勃发展
• 你不知道该用哪个

AI-assisted development tools are exploding:

• Claude Code, Cursor, Continue.dev, Aider...
• Each tool has its own commands and skills
• Skill packs like superpowers, gstack, omx are booming
• You don't know which one to use

解决方案 The Solution

# Just say what you want (自然语言输入)
vibe route "debug this database error"
# → Routes to: systematic-debugging (95% confidence)

vibe route "帮我扫描安全漏洞"
# → Routes to: gstack/cso (88% confidence)

vibe route "review my PR"
# → Routes to: gstack/review (92% confidence)

VibeSOP:

1. 理解你的意图 (自然语言，支持中英文)
2. 找到正确的技能 (从 45+ 技能中选择)
3. 学习你的偏好 (越用越准确)
4. 跨平台通用 (Claude Code, Cursor, Continue.dev 等)

VibeSOP:

1. Understands your intent (natural language, English + Chinese)
2. Finds the right skill (from 45+ available skills)
3. Learns your preferences (gets better over time)
4. Works with any AI tool (Claude Code, Cursor, Continue.dev, etc.)

---

核心功能 Core Features

🚀 一键智能安装（One-Click Smart Installation）

从 8 步手动配置 → 1 条命令，AI 自动完成所有配置

# 安装任何技能，零配置
vibe skill add tushare
vibe skill add git-helper
vibe skill add code-reviewer

# 系统自动完成：
# ✅ 检测技能类型和元数据
# ✅ 运行安全审计
# ✅ 询问安装范围（项目/全局）
# ✅ AI 智能配置路由规则
# ✅ 自动计算优先级
# ✅ 验证和同步到平台

对比传统方式：

特性	传统方式	VibeSOP
安装步骤	8+ 手动步骤	1 条命令
时间成本	30-60 分钟	1-2 分钟
配置文件	3-4 个手动编辑	0 个（AI 生成）
错误率	40%	<5%
学习曲线	陡峭	平缓

智能特性：

• 🤖 AI 配置引擎 - 分析技能描述，自动生成最优配置
• 🎯 智能路由 - 提取关键词，自动生成正则表达式
• ⚡ 优先级计算 - 基于技能类别自动设定优先级
• 🔒 安全审计 - 自动扫描，风险分级，交互式确认
• 💬 友好向导 - 清晰的进度展示和错误提示

完整文档 | .skill 格式规范

---

🎯 95% 路由准确率 (95% Routing Accuracy)

基于 10 层路由 pipeline，结合 AI 语义分析和场景知识：

Based on a 10-layer routing pipeline combining AI semantic analysis and scenario knowledge:

• Layer 0: Explicit overrides
• Layer 1: Scenario patterns (90% accuracy)
• Layer 2: AI Semantic Triage (95% accuracy, forced for long queries >5 chars)
• Layer 3: Keyword matching (70% accuracy, short queries only by default)
• Layer 4: TF-IDF semantic similarity (75% accuracy)
• Layer 5: Embedding-based matching (85% accuracy)
• Layer 6: Fuzzy matching for typos (60% accuracy)
• Layer 7: Custom plugin matchers (user-defined)
• Layer 8: No Match (below threshold)
• Layer 9: Fallback LLM (last-resort routing)

🛒 技能市场 (Skill Market) — v5.2.0

发现、安装、发布技能：

# 搜索市场中已有的技能
vibe market search "debug"

# 从市场安装技能
vibe market install user/repo

# 发布你的技能到市场（通过 GitHub Issue）
vibe market publish
vibe market publish --dry-run  # 预览发布内容

发布基于 GitHub Issues，无需额外服务器。关闭 Issue 即下架。

📉 智能降级 (Degradation) — v5.2.0

4 级置信度降级，替代二元 fallback：

>= 0.6 → 自动选择    (AUTO)
>= 0.4 → 建议确认    (SUGGEST)
>= 0.2 → 降级警告    (DEGRADE)
< 0.2  → 原始 LLM   (FALLBACK)

所有阈值可配置。用户显式指定的技能不受降级影响。

🔍 主动发现 (Proactive Discovery) — v5.2.0

每次路由后自动推荐尚未使用但匹配当前工作流的技能，标记为 [DISCOVER]。让你持续发现生态中适合你的技能。

🧠 偏好学习 (Preference Learning)

VibeSOP 会记住你的选择：

VibeSOP remembers your choices:

# First time
$ vibe route "debug this"
→ systematic-debugging (85%)

# You use it and it works
$ vibe feedback record "debug this" "systematic-debugging" --correct

# Next time
$ vibe route "debug this"
→ systematic-debugging (92%) ← Boosted!

🔓 开放生态 (Open Ecosystem)

不绑定任何平台，支持所有 AI 工具：

No platform lock-in, works with all AI tools:

• ✅ Claude Code
• ✅ Cursor
• ✅ Continue.dev
• ✅ Aider
• ✅ Any tool that supports SKILL.md

🛡️ 安全审计 (Security Audit)

每个外部技能都会经过安全扫描：

Every external skill is security-scanned:

• ✅ Prompt injection detection
• ✅ Command injection detection
• ✅ Role hijacking detection
• ✅ Privilege escalation detection
• ✅ Path traversal protection

---

使用示例 Usage Examples

调试错误 Debugging Errors

$ vibe route "database connection failed after deployment"

✅ Matched: systematic-debugging
   Rationale: Error detected → Use debugging workflow

代码审查 Code Review

$ vibe route "review my changes before pushing"

✅ Matched: gstack/review
   Confidence: 93%

中文查询 Chinese Queries

$ vibe route "帮我重构这个函数"

✅ Matched: superpowers/refactor
   Confidence: 89%

$ vibe route "代码覆盖率太低怎么办"

✅ Matched: superpowers/tdd
   Confidence: 91%

头脑风暴 Brainstorming

$ vibe route "I need ideas for a new feature"

✅ Matched: gstack/office-hours
   Confidence: 87%
   Rationale: "ideas" + "new feature" → design thinking

---

谁应该使用 VibeSOP？ Who Should Use VibeSOP?

👨‍💻 开发者 Developers

你正在使用 AI 辅助开发工具，但：

You're using AI-assisted development tools, but:

• ❌ 记不住那么多命令 / Can't remember all the commands
• ❌ 不知道哪个技能最适合当前场景 / Don't know which skill fits the current scenario
• ❌ 想要在不同工具间切换而不失去技能 / Want to switch tools without losing skills

VibeSOP 为你解决这些问题！ VibeSOP solves these problems for you!

🏢 团队 Teams

你们正在采用 AI 辅助开发，但：

You're adopting AI-assisted development, but:

• ❌ 团队成员使用不同的技能 / Team members use different skills
• ❌ 缺乏统一的技能管理 / Lack unified skill management
• ❌ 难以跟踪和分享最佳实践 / Hard to track and share best practices

VibeSOP 提供统一的技能管理和路由！ VibeSOP provides unified skill management and routing!

🌐 开源社区 Open Source Community

你正在维护 AI 辅助开发工具，但：

You're maintaining AI-assisted development tools, but:

• ❌ 技能格式不统一 / Inconsistent skill formats
• ❌ 难以集成外部技能 / Hard to integrate external skills
• ❌ 缺乏跨平台支持 / Lack cross-platform support

VibeSOP 提供标准的 SKILL.md 格式和跨平台支持！ VibeSOP provides standard SKILL.md format and cross-platform support!

---

CLI 命令参考 CLI Reference

核心命令 Core Commands

# Route query to best skill
vibe route "<query>"

# Orchestrate complex multi-intent query
vibe orchestrate "<query>"

# Decompose query into sub-tasks (without routing)
vibe decompose "<query>"

# List all available skills
vibe skills available

# Show skill details
vibe skills info <skill-id>

# Install skill pack
vibe install <url-or-name>

# Check environment
vibe doctor

技能管理 Skills Management

# List installed skills
vibe skills list

# Show detailed skill information
vibe skills info <skill-id>

# Install from URL or name
vibe install gstack
vibe install https://github.com/user/skills

# Sync skills to platform
vibe skills sync claude-code

反馈收集 Feedback Collection

# Record correct routing
vibe feedback record "<query>" "<skill>" --correct

# Record incorrect routing
vibe feedback record "<query>" "<skill>" --wrong "<actual-skill>"

# View feedback report
vibe feedback report

会话智能路由 Session Intelligent Routing

⚠️ Opt-in 设计：会话智能追踪默认关闭（VIBESOP_CONTEXT_TRACKING=false），这是有意的设计选择。

为什么默认关闭？

• 性能：零开销，不影响正常使用
• 隐私：不记录工具使用历史
• 控制：完全由用户决定是否启用

需要手动启用才能使用此功能。

# Enable tracking (Claude Code)
vibe session enable-tracking

# Record tool usage (manual)
vibe session record-tool --tool "read" --skill "systematic-debugging"

# Check for re-routing suggestions
vibe session check-reroute "design new architecture" --skill "systematic-debugging"

# View session summary
vibe session summary

完整命令参考: docs/user/CLI_REFERENCE.md Full CLI reference: docs/user/CLI_REFERENCE.md

---

配置 Configuration

项目级配置 Project-Level Config

创建 .vibe/config.yaml：

Create .vibe/config.yaml:

# .vibe/config.yaml
platform: claude-code

routing:
  min_confidence: 0.6
  enable_ai_triage: true
  enable_embedding: false
  max_candidates: 3
  confirmation_mode: always  # always | never | ambiguous_only
  keyword_match_max_chars: 5  # max chars for keyword routing (0=always LLM, 200=always keyword)

  # Degradation: confidence-gated layered fallback (v5.2.0)
  degradation_enabled: true
  degradation_auto_threshold: 0.6    # >= this = auto-select
  degradation_suggest_threshold: 0.4 # >= this but < auto = suggest
  degradation_degrade_threshold: 0.2 # >= this but < suggest = degrade
  degradation_fallback_always_ask: true  # ask user before raw LLM

security:
  threat_level: medium
  scan_external: true

skills:
  namespaces:
    - builtin
    - gstack
    - superpowers

用户确认模式 User Confirmation Mode

默认情况下，VibeSOP 会在选择技能前展示路由决策报告并要求你确认：

$ vibe route "帮我 review 代码"
╭────────── 🔍 Routing Decision Report ──────────╮
│ Selected: gstack/review (confidence: 87%)      │
│ ...                                            │
╰────────────────────────────────────────────────╯
How would you like to proceed?
  ✅ Confirm selected skill
  🔀 Choose a different skill
  📝 Skip skill, use raw LLM

你可以通过以下方式关闭确认：

• 临时跳过：vibe route "query" --yes 或 -y
• 全局关闭：在 ~/.vibe/config.yaml 中设置 routing.confirmation_mode: never
• 仅低置信度时确认：设置 routing.confirmation_mode: ambiguous_only

⚠️ 注意：确认模式默认开启 (always)，旨在让你了解 VibeSOP 的决策过程。关闭后将恢复为自动选择。

全局配置 Global Config

创建 ~/.vibe/config.yaml：

Create ~/.vibe/config.yaml:

# ~/.vibe/config.yaml
default_platform: claude-code
llm_provider: anthropic  # or openai

routing:
  enable_ai_triage: true
  use_cache: true

preferences:
  learning_enabled: true

---

集成 Integrations

Claude Code

# Build and deploy to Claude Code
vibe build claude-code --output ~/.claude

# Claude Code will now use VibeSOP for routing

Cursor

# Build for Cursor
vibe build cursor --output ~/.cursor

# Skills available in Cursor sessions

Continue.dev

# Build for Continue
vibe build opencode --output ~/.continue

# Use in Continue.dev configurations

---

架构 Architecture

┌─────────────────────────────────────────────────┐
│               AI Agent (执行层)                  │
│    Claude Code / Cursor / OpenCode / etc.        │
└────────────────────┬────────────────────────────┘
                     │ 执行技能
┌────────────────────▼────────────────────────────┐
│              VibeSOP SkillOS                     │
│                                                  │
│  ┌──────────────────────────────────────────┐   │
│  │         CLI / Agent Runtime Layer         │   │
│  │   vibe route │ orchestrate │ skill mgmt   │   │
│  └────────────────────┬─────────────────────┘   │
│                       │                         │
│  ┌────────────────────▼─────────────────────┐   │
│  │          UnifiedRouter (路由层)           │   │
│  │   10-Layer Pipeline:                      │   │
│  │   AI Triage → Scenario → Keyword → TF-IDF │   │
│  │   → Embedding → Levenshtein → Fallback    │   │
│  └────────────────────┬─────────────────────┘   │
│                       │                         │
│  ┌────────────────────▼─────────────────────┐   │
│  │       TaskOrchestrator (编排层)           │   │
│  │   多意图检测 → 任务分解 → 执行计划生成    │   │
│  └────────────────────┬─────────────────────┘   │
│                       │                         │
│  ┌────────────────────▼─────────────────────┐   │
│  │      Skill Lifecycle Manager (管理层)      │   │
│  │   启禁用 │ 作用域 │ 质量评估 │ 保留淘汰   │   │
│  └────────────────────┬─────────────────────┘   │
│                       │                         │
│  ┌────────────────────▼─────────────────────┐   │
│  │        Integration Layer (适配层)          │   │
│  │   Claude Code │ OpenCode │ Kimi CLI       │   │
│  └──────────────────────────────────────────┘   │
└─────────────────────────────────────────────────┘

详细架构文档: docs/architecture/ Detailed architecture docs: docs/architecture/

---

文档 Documentation

📚 完整文档索引 Complete Documentation Index: docs/INDEX.md

核心文档 Core Documentation

• docs/PHILOSOPHY.md - 核心哲学和使命 / Core philosophy and mission
• docs/architecture/ARCHITECTURE.md - 系统架构 / System architecture
• docs/PROJECT_CONTEXT.md - 项目背景 / Project context
• docs/PROJECT_STATUS.md - 项目状态 / Project status

用户指南 User Guides

• 🆕 docs/SKILLS_GUIDE.md - 技能生态系统完整指南 / Complete skills ecosystem guide
  • 50+ 个技能详解 / All skills explained
  • 10 层路由系统 / 10-layer routing system
  • 优先级决策机制 / Priority decision mechanism
  • 手动切换技能 / How to switch skills
• docs/QUICKSTART_USERS.md - 用户快速入门 / User quick start
• docs/QUICKSTART_DEVELOPERS.md - 开发者快速入门 / Developer quick start
• docs/user/CLI_REFERENCE.md - CLI 命令参考 / CLI command reference
• docs/EXTERNAL_SKILLS_GUIDE.md - 外部技能开发 / External skill development

技能包指南 Skill Pack Guides

• docs/OMX_GUIDE.md - oh-my-codex (OMX) 完整指南 / Complete OMX guide
  • deep-interview, ralph, ralplan, team, ultrawork, autopilot, ultraqa
  • 使用场景和最佳实践 / Usage scenarios and best practices

开发者文档 Developer Documentation

• docs/dev/CONTRIBUTING.md - 贡献指南 / Contributing guide
• docs/ROADMAP.md - 路线图 / Roadmap
• docs/CODE_OF_CONDUCT.md - 行为准则 / Code of conduct
• docs/SECURITY.md - 安全政策 / Security policy

---

性能指标 Performance Metrics

路由准确率 Routing Accuracy

指标 Metric	值 Value
总体准确率 Overall Accuracy	94%
AI Triage 准确率 AI Triage Accuracy	95%
场景匹配准确率 Scenario Matching Accuracy	90%
语义歧义准确率 Semantic Ambiguity Accuracy	90%

响应时间 Response Time

操作 Operation	时间 Time
简单路由 Simple Routing (缓存命中)	~10ms
复杂路由 Complex Routing (多层)	~270ms
AI Triage	~220ms

详见: docs/benchmarks/routing-accuracy-benchmark.md See: docs/benchmarks/routing-accuracy-benchmark.md

---

对比 Comparison

与其他工具对比 vs Other Tools

Feature	VibeSOP	Cursor	Continue.dev	Aider
Routing	10-layer intelligent routing	Built-in commands	Extension-based	CLI flags
Orchestration	Multi-skill composition	No	No	No
Lifecycle Mgmt	Enable/disable, scope, evaluate	No	No	No
Skills	50+ cross-platform skills	Built-in features	Community extensions	Built-in workflows
Learning	Preference learning	Fixed	No	No
Cross-Platform	✅ Works with any AI tool	❌ Cursor only	❌ Continue only	❌ Aider only
Open Ecosystem	✅ Any SKILL.md	❌ Closed	⚠️ Extension API	❌ Closed
Security Audit	✅ Before loading skills	N/A	⚠️ User discretion	N/A

为什么选择 VibeSOP？ Why Choose VibeSOP?

1. 不绑定单一工具 — 从 Cursor 切换到 Claude Code？你的技能跟着你走 Not tied to one tool — Switch from Cursor to Claude Code? Your skills come with you
2. 发现你不知道存在的技能 — "我能做什么？" → vibe skills available Discovers skills you didn't know existed — "What can I do?" → vibe skills available
3. 越来越聪明 — 记住什么对你有效 Gets smarter over time — Remembers what worked for you
4. 开放可扩展 — 用简单的 markdown 文件创建自己的技能 Open & extensible — Create your own skills with a simple markdown file

---

开发 Development

# Type checking
uv run basedpyright

# Linting
uv run ruff check

# Formatting
uv run ruff format

# Testing (fast, parallel, ~30s)
make test-fast

# Full test suite with coverage (~4 min)
uv run pytest

# Test coverage
uv run pytest --cov=src/vibesop --cov-report=html

---

路线图 Roadmap

• v4.0.0: 核心路由引擎 Core routing engine with 10-layer pipeline
• v4.1.0: AI Triage 生产就绪 AI Triage production readiness
• v4.2.0: 技能健康监控 Skill health monitoring
• v4.3.0: 上下文感知路由 + Agent Runtime Context-aware routing + Agent Runtime
• v4.4.0: SkillOS 编排 + 生命周期 + 反馈闭环 SkillOS Orchestration + Lifecycle + Feedback Loop
• v5.0.0: SkillRuntime — 作用域 + 生命周期 + 启禁用（质量收敛）
• v5.1.0: 技能市场 + 反馈闭环 SkillMarket + Feedback Loop
• v5.2.0: 智能生态系统 Intelligent Ecosystem — 推荐 + 退化 + 发现
• v5.3.0: 产品体验重塑 Product Experience — 仪表盘 + 清理 + 社区 + 徽章 + 引导

详见: docs/ROADMAP.md | version_05.md ADR

---

许可证 License

MIT License - see LICENSE file.

---

致谢 Acknowledgments

VibeSOP 站在巨人的肩膀上，整合了社区优秀的 AI 工程实践：

VibeSOP stands on the shoulders of giants, integrating excellent AI engineering practices from the community:

🔗 社区项目集成 Community Integration

VibeSOP 内置了对以下社区技能包的支持，并提供统一的智能路由：

VibeSOP provides built-in support and intelligent routing for the following community skill packs:

• gstack by @brandonrobertz

  • 🎯 定位: 虚拟工程团队 - 工程技能和浏览器自动化
  • 📦 技能数: 19 个技能 (review, qa, ship, office-hours, browse, etc.)
  • 🎨 特点: 角色-based 技能 (产品、工程、设计、QA)
  • 📖 文档: OMX_GUIDE.md (参见对比章节)

• superpowers by @obra

  • 🎯 定位: 基础开发工作流 - TDD、重构、调试、优化
  • 📦 技能数: 7 个技能 (tdd, refactor, debug, optimize, architect, review, brainstorm)
  • 🎨 特点: 开发最佳实践，red-green-refactor 循环
  • 💡 适用: 日常开发任务，个人工作流优化

• oh-my-codex (OMX) by @mill173

  • 🎯 定位: 高级工程方法论 - 结构化思维和系统化执行
  • 📦 技能数: 7 个技能 (deep-interview, ralph, ralplan, team, ultrawork, autopilot, ultraqa)
  • 🎨 特点: 需求澄清、持久执行、共识规划、多代理并行
  • 📖 文档: OMX_GUIDE.md (完整使用指南)

🏗️ 核心技术基础 Core Technologies

• Claude Code by Anthropic
  • 📋 贡献: SKILL.md 规范标准
  • 🔧 集成: VibeSOP 完全兼容 SKILL.md 规范
  • 📚 文档: SKILL.md Specification

🎯 VibeSOP 独特价值 VibeSOP Unique Value

VibeSOP 不仅仅是这些技能包的集合，而是一个统一的智能路由层：

VibeSOP is not just a collection of these skill packs, but a unified intelligent routing layer:

• 🧠 智能路由 (94% 准确率) - 自动选择最合适的技能
• 🔄 统一管理 - 一个工具管理所有技能包
• 🛡️ 安全审计 - 所有外部技能经过安全扫描
• 📚 跨平台 - 在 Claude Code、Cursor、Continue.dev 等平台使用
• 🎓 偏好学习 - 记住你的选择，越来越准确

📊 技能选择指南 Skill Selection Guide

详细对比: 请参考 OMX_GUIDE.md

需求不明确？ → OMX deep-interview (深度澄清)
产品头脑风暴？ → GStack office-hours (重新框定)
TDD 开发？ → Superpowers tdd (red-green-refactor)
代码审查？ → GStack review (SQL 安全 + 自动修复)
调试错误？ → Systematic debugging (内置，根因分析)
完整实现？ → OMX ralph (持久执行 + deslop)
团队决策？ → OMX ralplan (共识规划 + ADR)
并行任务？ → OMX team (多代理协作)
QA 测试？ → GStack qa (浏览器自动化) 或 OMX ultraqa (架构驱动)
准备发布？ → GStack ship (完整发布流程)

💬 致谢社区 Thanks to the Community

感谢以下项目的作者和维护者，他们的工作让 AI 原生开发更加强大：

Thanks to the authors and maintainers of these projects for making AI-native development more powerful:

• @brandonrobertz - gstack
• @obra - superpowers
• @mill173 - oh-my-codex (OMX)
• Anthropic Team - Claude Code

---

🆕 智能技能安装特性

⚡ 一键安装，零配置

从 8 步手动配置 → 1 条命令，98% 时间节省

# 安装任何技能
vibe skill add tushare
vibe skill add git-helper
vibe skill add code-reviewer

# AI 自动完成：
# ✅ 检测技能类型和元数据
# ✅ 安全审计（自动扫描，风险分级）
# ✅ 智能配置（路由规则、优先级、标签）
# ✅ 验证和同步（自动测试，同步平台）

📊 性能对比

指标	传统方式	VibeSOP	改进
安装步骤	8+ 手动步骤	1 条命令	87.5% ↓
时间成本	30-60 分钟	1-2 分钟	95% ↓
配置文件	3-4 个手动编辑	0 个（AI 生成）	100% ↓
出错率	40%	<5%	87.5% ↓
满意度	2.5/5	4.8/5	92% ↑

🎯 核心特性

• 🤖 AI 智能配置 - 分析技能描述，自动生成最优配置
• 🎯 智能路由 - 提取关键词，自动生成正则表达式
• ⚡ 优先级计算 - 基于技能类别自动设定优先级
• 🔒 安全审计 - 自动扫描，风险分级，交互式确认
• 💬 友好向导 - 清晰的进度展示和错误提示
• 📦 标准格式 - .skill 统一分发和安装格式

📖 完整文档 | .skill 规范

---

联系我们 Contact Us

用 ❤️ 构建，为 AI 原生开发工作流 Built with ❤️ for AI-native developer workflows

GitHub • Issues • Discussions

---

版本 Version: 5.3.0 更新时间 Last Updated: 2026-04-28 状态 Status: ✅ 生产就绪 Production Ready