sumulige-claude 1.5.2 → 1.6.0

package/README.md

@@ -1,682 +1,433 @@
-# Sumulige Claude
+<p align="center">
+  <h1 align="center">Sumulige Claude</h1>
+  <p align="center">
+    <strong>The Universal Agent Harness for AI Coding Assistants</strong>
+  </p>
+  <p align="center">
+    Transform Claude Code and OpenAI Codex into intelligent, memory-aware development teams
+  </p>
+</p>
+
+<p align="center">
+  <a href="https://www.npmjs.com/package/sumulige-claude"><img src="https://badge.fury.io/js/sumulige-claude.svg" alt="npm version"></a>
+  <a href="https://opensource.org/licenses/MIT"><img src="https://img.shields.io/badge/License-MIT-yellow.svg" alt="License: MIT"></a>
+  <a href="#"><img src="https://img.shields.io/badge/platforms-Claude%20%7C%20Codex-blue.svg" alt="Platforms"></a>
+</p>
 
-**The Best Agent Harness for Claude Code**
-**Claude Code 的终极 Agent 编排框架**
+---
 
-[![npm version](https://badge.fury.io/js/sumulige-claude.svg)](https://www.npmjs.com/package/sumulige-claude)
-[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
+## Quick Start
 
----
+```bash
+# Install
+npm install -g sumulige-claude
 
-## Table of Contents / 目录
+# Initialize in your project
+smc template
+
+# Start Claude Code
+claude
+```
 
-1. [Layer 1: What is it? / 第一层：它是什么？](#layer-1-what-is-it--第一层它是什么)
-2. [Layer 2: Why use it? / 第二层：为什么使用它？](#layer-2-why-use-it--第二层为什么使用它)
-3. [Layer 3: Core Concepts / 第三层：核心概念](#layer-3-core-concepts--第三层核心概念)
-4. [Layer 4: Quick Start / 第四层：快速开始](#layer-4-quick-start--第四层快速开始)
-5. [Layer 5: Commands / 第五层：命令参考](#layer-5-commands--第五层命令参考)
-6. [Layer 6: Advanced / 第六层：高级配置](#layer-6-advanced--第六层高级配置)
-7. [Layer 7: Lifecycle Hooks / 第七层：生命周期钩子](#layer-7-lifecycle-hooks--第七层生命周期钩子)
-8. [Layer 8: Data Flow Architecture / 第八层：数据流转架构](#layer-8-data-flow-architecture--第八层数据流转架构)
-9. [Layer 9: Config Inheritance / 第九层：配置分层继承](#layer-9-config-inheritance--第九层配置分层继承)
-10. [Documentation / 文档](#documentation--文档)
+**That's it.** Your project now has AI memory, slash commands, and quality gates.
 
 ---
 
-## Layer 1: What is it? / 第一层：它是什么？
+## Why Sumulige Claude?
 
-### The 30-Second Version / 30 秒版本
+| Problem | Before | After |
+|---------|--------|-------|
+| AI forgets context every session | Repeat project structure constantly | Automatic memory via ThinkingLens |
+| Inconsistent code quality | Manual reviews, missed issues | Quality Gate auto-checks |
+| Works with Claude, need Codex too | Maintain two config systems | One config, both platforms |
 
-**English** | Sumulige Claude is a CLI tool that makes Claude Code smarter by:
-- Adding specialized AI agents (like a team of experts)
-- Providing ready-to-use project templates
-- Managing reusable "skills" for common tasks
+---
 
-**中文** | Sumulige Claude 是一个让 Claude Code 更聪明的 CLI 工具：
-- 添加专业的 AI Agent（像专家团队）
-- 提供开箱即用的项目模板
-- 管理可复用的"技能"
+## Features
 
-### The 5-Minute Version / 5 分钟版本
+### Multi-Platform Support
 
 ```
 ┌─────────────────────────────────────────────────────────────┐
-│                    Without Sumulige Claude                  │
+│                    smc CLI                                   │
 ├─────────────────────────────────────────────────────────────┤
-│  你 ──▶ Claude ──▶ 写代码                                    │
-│  │                                                          │
-│  └─ 每次都要解释项目结构                                     │
-│  └─ 要重复告诉 Claude 代码风格                               │
-│  └─ Claude 不知道你的团队约定                                │
-└─────────────────────────────────────────────────────────────┘
-
-┌─────────────────────────────────────────────────────────────┐
-│                    With Sumulige Claude                     │
+│  ┌─────────────────┐          ┌─────────────────┐           │
+│  │   Claude Code   │          │   Codex CLI     │           │
+│  │   (Anthropic)   │          │   (OpenAI)      │           │
+│  │                 │          │                 │           │
+│  │  .claude/       │  ◄────►  │  .codex/        │           │
+│  │  CLAUDE.md      │  sync    │  AGENTS.md      │           │
+│  │  JSON config    │          │  TOML config    │           │
+│  └─────────────────┘          └─────────────────┘           │
 ├─────────────────────────────────────────────────────────────┤
-│  你 ──▶ Conductor ──▶ Architect ──▶ Builder ──▶ 代码        │
-│       (协调)        (架构)         (实现)                    │
-│                                                             │
-│  ✓ 自动理解项目结构                                          │
-│  ✓ 知道团队代码风格                                          │
-│  ✓ 记住所有历史决策                                          │
+│              Shared: Skills, Rules, Memory                   │
 └─────────────────────────────────────────────────────────────┘
 ```
 
+### Core Capabilities
+
+- **Memory System** - AI remembers decisions across sessions
+- **Multi-Agent** - Specialized AI agents (Architect, Builder, Reviewer)
+- **Skills Marketplace** - Install and share reusable AI capabilities
+- **Quality Gate** - Automatic code quality enforcement
+- **Slash Commands** - `/commit`, `/test`, `/review`, `/fix`
+
 ---
 
-## Layer 2: Why use it? / 第二层：为什么使用它？
+## Installation
 
-### Problem vs Solution / 问题 vs 解决方案
+### Requirements
 
-| 问题 | 传统方式 | Sumulige Claude |
-|------|---------|-----------------|
-| **每次都要解释项目结构** | 复制粘贴 README 内容 | `.claude/rag/` 自动存储项目知识 |
-| **代码风格不一致** | 手动告诉 Claude 规则 | Quality Gate 自动检查 |
-| **AI 不记得之前的决策** | 翻聊天记录找答案 | ThinkingLens 记忆所有对话 |
-| **团队成员配置不同** | 各自配置 Claude | `smc sync` 统一环境 |
-| **重复性任务** | 每次重新写 Prompt | Skills 系统，一键调用 |
+- Node.js 16+
+- Claude Code or Codex CLI installed
 
-### Three Key Benefits / 三大核心价值
+### Install
 
+```bash
+npm install -g sumulige-claude
 ```
-1. 记忆 (Memory)
-   ┌─────────────────────────────────────────────┐
-   │  ThinkingLens                                │
-   │  ├─ 记住每次对话的上下文                      │
-   │  ├─ 记住项目的架构决策                        │
-   │  └─ 记住代码风格约定                          │
-   └─────────────────────────────────────────────┘
-
-2. 分工 (Specialization)
-   ┌─────────────────────────────────────────────┐
-   │  Multi-Agent System                          │
-   │  ├─ Conductor  → 任务分解                    │
-   │  ├─ Architect  → 架构设计                    │
-   │  ├─ Builder    → 代码实现                    │
-   │  ├─ Reviewer   → 质量检查                    │
-   │  └─ Librarian  → 文档整理                    │
-   └─────────────────────────────────────────────┘
-
-3. 复用 (Reuse)
-   ┌─────────────────────────────────────────────┐
-   │  Skills System                               │
-   │  ├─ 安装社区技能                             │
-   │  ├─ 创建自己的技能                           │
-   │  └─ 团队共享技能库                           │
-   └─────────────────────────────────────────────┘
+
+### Verify
+
+```bash
+smc --version
 ```
 
 ---
 
-## Layer 3: Core Concepts / 第三层：核心概念
+## Usage
 
-### Concept 1: What is an "Agent"? / 什么是 "Agent"？
+### New Project Setup
 
-**English** | An Agent is like a virtual team member with a specific role. Instead of one AI doing everything, you have specialized AIs working together.
+```bash
+# Deploy Claude Code template
+smc template
 
-**中文** | Agent 就像是一个有特定职责的虚拟团队成员。不是让一个 AI 做所有事，而是让专业的 AI 协作。
+# Or deploy for both platforms
+smc template --all
 
+# Or Codex only
+smc template --codex
 ```
-传统方式：
-┌─────────────────────────────────────┐
-│         Claude (通用 AI)             │
-│  写代码 + 审查 + 架构 + 文档         │
-└─────────────────────────────────────┘
-
-Agent 方式：
-┌─────────┐  ┌─────────┐  ┌─────────┐
-│Conductor│─▶│Architect│─▶│ Builder │
-│ 分解任务 │  │ 设计架构 │  │ 实现代码 │
-└─────────┘  └─────────┘  └─────────┘
-                    │
-                    ▼
-              ┌─────────┐
-              │Reviewer │
-              │ 审查代码 │
-              └─────────┘
-```
-
-### Concept 2: What is a "Skill"? / 什么是 "Skill"？
 
-**English** | A Skill is a reusable set of instructions for Claude. Think of it as a "plugin" that gives Claude new capabilities.
+### Existing Project
 
-**中文** | Skill 是一组可复用的指令，就像是给 Claude 安装"插件"，赋予它新的能力。
+```bash
+# Add to existing project (safe mode - won't overwrite)
+smc template --safe
 
-```
-Skill = 上下文 + 指令 + 资源
-
-┌─────────────────────────────────────┐
-│  Skill: api-tester                  │
-├─────────────────────────────────────┤
-│  上下文: REST/GraphQL API 测试知识   │
-│  指令: 如何验证 API 响应             │
-│  资源: 测试脚本模板                  │
-└─────────────────────────────────────┘
+# Or sync incrementally
+smc sync
 ```
 
-### Concept 3: What is the Template? / 什么是模板？
+### Platform Commands
 
-**English** | The Template is a pre-configured project structure with all AI settings, hooks, and skills ready to use.
+```bash
+# Detect configured platforms
+smc platform:detect
 
-**中文** | 模板是预先配置好的项目结构，包含所有 AI 设置、钩子和技能，开箱即用。
+# Convert Claude config to Codex
+smc platform:convert claude codex
 
-```
-smc template 之后，你的项目会变成：
-
-your-project/
-├── .claude/
-│   ├── commands/      # 斜杠命令定义
-│   ├── skills/        # 技能库
-│   ├── hooks/         # 自动化钩子
-│   ├── rag/           # 项目知识库
-│   └── sessions/      # 对话历史
-├── CLAUDE.md          # AI 配置文件
-└── development/
-    └── todos/         # 任务管理系统
+# Sync to all platforms
+smc platform:sync
 ```
 
 ---
 
-## Layer 4: Quick Start / 第四层：快速开始
+## Project Structure
 
-### Installation / 安装
+After `smc template`, your project looks like:
 
-```bash
-npm install -g sumulige-claude
+```
+your-project/
+├── .claude/                 # Claude Code configuration
+│   ├── settings.json        # Hooks and settings
+│   ├── CLAUDE.md            # Project instructions
+│   ├── commands/            # Slash commands
+│   ├── skills/              # Installed skills
+│   ├── hooks/               # Automation hooks
+│   ├── rag/                 # Knowledge index
+│   └── MEMORY.md            # AI memory
+│
+├── .codex/                  # Codex CLI configuration (if --all)
+│   └── config.toml          # Codex settings
+│
+├── AGENTS.md                # Codex instructions (if --all)
+└── CLAUDE.md                # Project-level AI config
 ```
 
-### Three Commands to Get Started / 三个命令开始使用
+---
 
-```bash
-# Step 1: Initialize global config / 初始化全局配置
-smc init
+## Commands Reference
 
-# Step 2: Deploy to your project / 部署到项目
-mkdir my-project && cd my-project
-smc template
+### Core Commands
 
-# Step 3: Start planning / 开始规划
-smc kickoff
-```
+| Command | Description |
+|---------|-------------|
+| `smc init` | Initialize global configuration |
+| `smc template` | Deploy project template |
+| `smc sync` | Sync configuration (incremental) |
+| `smc status` | Show current status |
 
-### What Each Command Does / 每个命令做什么
+### Platform Commands
 
-| 命令 | 做什么 | 类比 |
-|------|--------|------|
-| `smc init` | 创建 `~/.claude/config.json` | 买房前准备工具箱 |
-| `smc template` | 复制模板到项目 | 装修新房子 |
-| `smc kickoff` | 启动项目规划 | 开工典礼 |
+| Command | Description |
+|---------|-------------|
+| `smc platform:detect` | Detect AI platforms in project |
+| `smc platform:list` | List supported platforms |
+| `smc platform:convert <from> <to>` | Convert between platforms |
+| `smc platform:sync` | Sync config to all platforms |
 
----
+### Skill Commands
 
-## Layer 5: Commands / 第五层：命令参考
+| Command | Description |
+|---------|-------------|
+| `smc skill:list` | List installed skills |
+| `smc skill:create <name>` | Create new skill |
+| `smc marketplace:list` | Browse skill marketplace |
+| `smc marketplace:install <name>` | Install a skill |
 
-### Command Groups / 命令分组
+### Quality Commands
 
-```
-┌─────────────────────────────────────────────────────────────┐
-│  基础命令 (每次使用都需要)                                    │
-├─────────────────────────────────────────────────────────────┤
-│  smc init          初始化全局配置                            │
-│  smc status        查看当前状态                              │
-│  smc sync          同步到当前项目（增量）                     │
-└─────────────────────────────────────────────────────────────┘
+| Command | Description |
+|---------|-------------|
+| `smc qg:check` | Run quality gate checks |
+| `smc qg:rules` | List quality rules |
+| `smc config:validate` | Validate configuration |
 
-┌─────────────────────────────────────────────────────────────┐
-│  项目命令 (新项目启动时使用)                                  │
-├─────────────────────────────────────────────────────────────┤
-│  smc template      部署项目模板                              │
-│  smc kickoff       启动项目规划                              │
-└─────────────────────────────────────────────────────────────┘
-
-┌─────────────────────────────────────────────────────────────┐
-│  技能命令 (管理 Claude 能力)                                 │
-├─────────────────────────────────────────────────────────────┤
-│  smc skill:list    列出已安装技能                            │
-│  smc marketplace:list  浏览技能市场                          │
-│  smc marketplace:install <name>  安装技能                   │
-└─────────────────────────────────────────────────────────────┘
+---
 
-┌─────────────────────────────────────────────────────────────┐
-│  配置命令 (高级配置)                                          │
-├─────────────────────────────────────────────────────────────┤
-│  smc config:validate   验证配置文件                          │
-│  smc config:backup     创建配置备份                          │
-│  smc qg:check          运行质量检查                          │
-└─────────────────────────────────────────────────────────────┘
-```
+## Slash Commands
 
-### Command Flow / 命令流程
+Available in Claude Code after template deployment:
 
-```
-新项目流程：
-┌─────────┐    ┌─────────┐    ┌─────────┐    ┌─────────┐
-│  新建   │──▶ │ smc     │──▶ │ smc     │──▶ │ smc     │
-│  项目   │    │ template │    │ kickoff │    │  开发   │
-└─────────┘    └─────────┘    └─────────┘    └─────────┘
-
-已有项目：
-┌─────────┐    ┌─────────┐    ┌─────────┐
-│ 进入    │──▶ │ smc     │──▶ │  开发   │
-│ 项目    │    │ sync    │    │         │
-└─────────┘    └─────────┘    └─────────┘
-```
+| Command | Model | Description |
+|---------|-------|-------------|
+| `/review` | Sonnet | Code review + security scan |
+| `/test` | Sonnet | Run tests with TDD support |
+| `/fix` | Haiku | Quick fix for build/lint errors |
+| `/plan` | Opus | Architecture and planning |
+| `/commit` | - | Git commit with message |
 
 ---
 
-## Layer 5.5: Core Skills / 核心技能系统
+## Multi-Platform Workflow
 
-### 4 Core Skills / 4 个核心技能
+### Converting Existing Project
 
-> v2.0.0 优化：将 9 个技能合并为 4 个核心技能，成本降低 60-70%
+```bash
+# You have a Claude Code project, want to add Codex support
 
-| Skill | 命令 | 模型 | 作用 |
-|-------|------|------|------|
-| `quality-guard` | `/review` | sonnet | 代码审查 + 安全检查 + 清理建议 |
-| `test-master` | `/test` | sonnet | TDD + E2E + 覆盖率分析 |
-| `design-brain` | `/plan` | opus | 规划 + 架构设计 |
-| `quick-fix` | `/fix` | haiku | 快速修复构建/类型/lint 错误 |
+# 1. See what's configured
+smc platform:detect
+# Output: Detected 1 platform(s): claude
 
-### Quick Reference / 命令速查
+# 2. Convert to Codex
+smc platform:convert claude codex
 
-```bash
-# 代码审查
-/review              # 标准审查
-/review --security   # 安全深扫 (OWASP Top 10)
-/review --clean      # 死代码清理
-
-# 测试
-/test                # 运行测试
-/test --tdd          # TDD 模式
-/test --e2e          # E2E 测试
-/test --coverage     # 覆盖率
-
-# 规划
-/plan                # 快速规划
-/plan --deep         # 深度设计
-
-# 快速修复
-/fix                 # 自动检测修复
-/fix --build         # 构建错误
-/fix --type          # 类型错误
-
-# 重构
-/refactor            # 分析建议
-/refactor --execute  # 执行清理
-```
+# 3. Verify
+smc platform:detect
+# Output: Detected 2 platform(s): claude, codex
 
-### Model Cost Strategy / 模型成本策略
-
-```
-┌─────────┬─────────────────────┬──────────┬────────┐
-│  模型   │      使用场景        │ 相对成本 │  速度  │
-├─────────┼─────────────────────┼──────────┼────────┤
-│ haiku   │ /fix 快速修复        │   1x     │  最快  │
-│ sonnet  │ /review, /test      │   5x     │  中等  │
-│ opus    │ /plan --deep        │  15x     │  较慢  │
-└─────────┴─────────────────────┴──────────┴────────┘
+# 4. Use either CLI
+claude "Fix the login bug"
+codex "Fix the login bug"
 ```
 
-详细使用指南: [.claude/USAGE.md](.claude/USAGE.md)
-
----
+### Platform Config Mapping
 
-## Layer 6: Advanced / 第六层：高级配置
+| Claude Code | Codex CLI |
+|-------------|-----------|
+| `.claude/settings.json` | `.codex/config.toml` |
+| `CLAUDE.md` | `AGENTS.md` |
+| JSON format | TOML format |
+| Hooks in settings.json | Notifications in config.toml |
 
-### Multi-Agent Configuration / 多 Agent 配置
+---
 
-**English** | Each agent can be customized with its own model and role description.
+## Configuration
 
-**中文** | 每个 Agent 都可以自定义模型和职责描述。
+### Claude Code (`~/.claude/config.json`)
 
 ```json
 {
-  "version": "1.2.0",
-  "model": "claude-opus-4.5",
+  "version": "1.6.0",
+  "model": "claude-sonnet-4",
   "agents": {
-    "conductor": {
-      "role": "任务协调与分解 - 理解需求并分配给合适的 Agent",
-      "model": "claude-opus-4.5"
-    },
-    "architect": {
-      "role": "架构设计 - 设计系统架构和技术选型",
-      "model": "claude-sonnet-4.5"
-    },
-    "builder": {
-      "role": "代码实现 - 编写高质量代码和测试",
-      "model": "claude-sonnet-4.5"
-    },
-    "reviewer": {
-      "role": "代码审查 - 检查代码质量和安全问题",
-      "model": "claude-opus-4.5"
-    },
-    "librarian": {
-      "role": "文档管理 - 整理文档和知识库",
-      "model": "claude-haiku-4.5"
-    }
+    "architect": { "model": "claude-opus-4" },
+    "builder": { "model": "claude-sonnet-4" },
+    "reviewer": { "model": "claude-opus-4" }
   },
-  "skills": [
-    "anthropics/skills",
-    "numman-ali/n-skills"
-  ],
   "thinkingLens": {
     "enabled": true,
-    "autoSync": true,
-    "syncInterval": 20
+    "autoSync": true
   }
 }
 ```
 
-### Quality Gate / 质量门禁
-
-**English** | Automatically check code quality before commits.
+### Codex CLI (`.codex/config.toml`)
 
-**中文** | 提交前自动检查代码质量。
-
-```json
-{
-  "enabled": true,
-  "severity": "warn",
-  "rules": [
-    { "id": "line-count-limit", "enabled": true },
-    { "id": "no-console-logs", "enabled": true }
-  ],
-  "gates": {
-    "preCommit": true,
-    "prePush": true
-  }
-}
-```
-
-### Skills Marketplace / 技能市场
-
-```bash
-# Browse available skills / 浏览可用技能
-smc marketplace:list
+```toml
+model = "o3"
+model_provider = "openai"
+sandbox_mode = "workspace-write"
+approval_policy = "on-failure"
 
-# Install a skill / 安装技能
-smc marketplace:install dev-browser
+[project]
+project_doc_fallback_filenames = ["AGENTS.md", "CLAUDE.md"]
 
-# Create your own skill / 创建自己的技能
-smc skill:create my-skill
+[features]
+shell_tool = true
+web_search_request = true
 ```
 
 ---
 
-## Layer 7: Lifecycle Hooks / 第七层：生命周期钩子
+## Architecture
 
-### Auto-Sync System / 自动同步系统
-
-> v1.3.2: 利用 Claude Code 官方 Hook 事件实现记忆自动同步
+### Memory System
 
 ```
-┌─────────────────────────────────────────────────────────────┐
-│  SessionStart ──► memory-loader.cjs                         │
-│    └─ 自动加载 MEMORY.md, ANCHORS.md, TODO 状态              │
-│                                                              │
-│  PreCompact ──► auto-handoff.cjs                            │
-│    └─ 上下文压缩前自动生成 handoff 文档                       │
-│                                                              │
-│  SessionEnd ──► memory-saver.cjs                            │
-│    └─ 会话结束自动保存摘要到 MEMORY.md                        │
-└─────────────────────────────────────────────────────────────┘
+Session Start
+     │
+     ▼
+┌─────────────────┐
+│ memory-loader   │ ◄── Load MEMORY.md, ANCHORS.md
+└─────────────────┘
+     │
+     ▼
+   Work with AI
+     │
+     ▼
+┌─────────────────┐
+│ memory-saver    │ ──► Save to MEMORY.md
+└─────────────────┘
+     │
+     ▼
+Session End
 ```
 
-### Hook Files / 钩子文件
-
-| Hook | 文件 | 触发时机 | 功能 |
-|------|------|----------|------|
-| `SessionStart` | `memory-loader.cjs` | 会话开始 | 加载记忆、锚点、TODO |
-| `SessionEnd` | `memory-saver.cjs` | 会话结束 | 保存摘要、归档会话 |
-| `PreCompact` | `auto-handoff.cjs` | 上下文压缩前 | 生成 handoff 保护上下文 |
-
-### Handoff Documents / 交接文档
-
-当上下文即将被压缩时，自动生成交接文档：
+### Hook System
 
 ```
-.claude/handoffs/
-├── LATEST.md           # 最新交接文档
-├── INDEX.md            # 交接文档索引
-└── handoff_*.md        # 历史交接文档
+Claude Code Event
+     │
+     ▼
+┌─────────────────────────────────────────┐
+│            settings.json                 │
+├─────────────────────────────────────────┤
+│ SessionStart  → memory-loader.cjs       │
+│ SessionEnd    → memory-saver.cjs        │
+│ PreCompact    → auto-handoff.cjs        │
+│ PostToolUse   → code-formatter.cjs      │
+└─────────────────────────────────────────┘
 ```
 
-每个交接文档包含：
-- 会话信息（项目、版本、开始时间）
-- 活跃 TODOs 列表
-- 最近修改的文件
-- 恢复命令
+---
 
-### Update Hooks / 更新钩子
+## Upgrading
 
-其他项目如何获取新 hooks：
+### From v1.x
 
 ```bash
-# 方式 1: 完整更新（推荐）
-smc template --force
-
-# 方式 2: 增量同步（仅更新 hooks）
-smc sync --hooks
-
-# 方式 3: 手动安装
+# Update global package
 npm update -g sumulige-claude
-smc template
-```
-
----
 
-## Layer 8: Data Flow Architecture / 第八层：数据流转架构
-
-> v1.4.0: Hook 统一调度 + 智能过滤，Token 成本降低 62%
-
-### Hook 调度流程
-
-```
-┌─────────────────────────────────────────────────────────────────────────┐
-│                           Claude Code 事件触发                            │
-└─────────────────────────────────────────────────────────────────────────┘
-                                    │
-                                    ▼
-┌─────────────────────────────────────────────────────────────────────────┐
-│                         settings.json 路由                               │
-├─────────────────────────────────────────────────────────────────────────┤
-│  SessionStart  ──────→ memory-loader.cjs (直接执行)                      │
-│  SessionEnd    ──────→ memory-saver.cjs (直接执行)                       │
-│  PreCompact    ──────→ auto-handoff.cjs (直接执行)                       │
-│  UserPromptSubmit ───→ hook-dispatcher.cjs                              │
-│  PreToolUse    ──────→ (空，不执行任何 hook)                              │
-│  PostToolUse   ──────→ code-formatter.cjs (仅 Write/Edit 时)            │
-│  AgentStop     ──────→ hook-dispatcher.cjs                              │
-└─────────────────────────────────────────────────────────────────────────┘
-```
-
-### Hook Dispatcher 内部流程
-
-```
-┌─────────────────────────────────────────────────────────────────────────┐
-│                      hook-dispatcher.cjs                                 │
-└─────────────────────────────────────────────────────────────────────────┘
-                                    │
-                    ┌───────────────┼───────────────┐
-                    ▼               ▼               ▼
-            ┌─────────────┐ ┌─────────────┐ ┌─────────────┐
-            │ 加载 Registry│ │  加载 State │ │ 获取事件类型 │
-            │ hook-registry│ │ .dispatcher │ │CLAUDE_EVENT │
-            │    .json    │ │ -state.json │ │   _TYPE     │
-            └─────────────┘ └─────────────┘ └─────────────┘
-                    │               │               │
-                    └───────────────┼───────────────┘
-                                    ▼
-                    ┌───────────────────────────────┐
-                    │      筛选当前事件的 Hooks      │
-                    └───────────────────────────────┘
-                                    │
-                                    ▼
-            ┌───────────────────────────────────────────────┐
-            │              对每个 Hook 检查                  │
-            ├───────────────────────────────────────────────┤
-            │  1. shouldRunDebounce() - 5秒内执行过? → 跳过  │
-            │  2. shouldRunOnce() - 已执行过? → 跳过         │
-            │  3. shouldRunCondition() - 条件不满足? → 跳过  │
-            └───────────────────────────────────────────────┘
-                                    │
-                                    ▼
-                    ┌───────────────────────────────┐
-                    │   执行通过检查的 Hook + 更新状态 │
-                    └───────────────────────────────┘
-```
-
-### 增量同步流程
-
-```
-┌─────────────────────────────────────────────────────────────────────────┐
-│                    smc sync --incremental                                │
-└─────────────────────────────────────────────────────────────────────────┘
-                                    │
-        ┌───────────────────────────┼───────────────────────────┐
-        ▼                           ▼                           ▼
-┌───────────────┐         ┌───────────────┐         ┌───────────────┐
-│ 读取项目版本   │         │ 读取版本清单   │         │  计算差异变更  │
-│ .sumulige-    │         │ version-      │         │ getChangesSince│
-│ claude-version│         │ manifest.json │         │   (1.3.3)     │
-└───────────────┘         └───────────────┘         └───────────────┘
-                                    │
-                                    ▼
-            ┌───────────────────────────────────────────────┐
-            │  应用变更: hook/config/lib/command            │
-            │  更新项目版本标记                              │
-            └───────────────────────────────────────────────┘
+# Update project template
+smc template --force
 ```
 
-### 状态文件汇总
+### Migration
 
-| 文件 | 用途 | 生命周期 |
-|------|------|----------|
-| `.dispatcher-state.json` | Hook 执行状态 | 会话级 |
-| `.match-cache.json` | RAG 匹配缓存 | 5 分钟 TTL |
-| `.sumulige-claude-version` | 项目版本标记 | 永久 |
+Old hooks are automatically migrated. Your custom configurations are backed up to `.claude/backup/`.
 
 ---
 
-## Layer 9: Config Inheritance / 第九层：配置分层继承
-
-> v1.4.1: 单一真相源架构，零维护成本的多项目配置管理
+## Troubleshooting
 
-### 架构图
+### Common Issues
 
-```
-sumulige-claude/.claude/          ~/.claude/              项目/.claude/
-(Git 仓库 - 版本控制)             (运行时 - 符号链接)      (项目特定)
-        │                              │                       │
-        ├── hooks/ ─────→ symlink ────→│                       │
-        ├── skills/ ────→ symlink ────→│        自动继承       │
-        ├── templates/ ─→ symlink ────→│ ─────────────────────→│
-        ├── commands/ ──→ symlink ────→│                       │
-        │                              │                       │
-        └── *.md ────────→ 复制 ──────→│                       │
-                                       │                       │
-                                       ├── cache/              ├── MEMORY.md
-                                       ├── debug/              ├── PROJECT_LOG.md
-                                       └── history.jsonl       └── thinking-routes/
+**Q: Commands not found after install**
+```bash
+# Ensure npm global bin is in PATH
+export PATH="$PATH:$(npm bin -g)"
 ```
 
-### 使用方式
-
+**Q: Platform not detected**
 ```bash
-# 首次设置：建立符号链接
-cd ~/Documents/Antigravity/sumulige-claude
-./scripts/sync-to-home.sh
-
-# 日常更新：修改 sumulige-claude 后自动生效
-git pull  # 所有项目自动获得最新配置！
-
-# 更新模板文件（CLAUDE.md, MEMORY.md 等）
-./scripts/sync-to-home.sh --copy
+# Ensure project has config files
+ls -la .claude/settings.json
+ls -la .codex/config.toml
 ```
 
-### 项目配置精简
-
-**之前**：每个项目复制完整 `.claude/` (~300 个文件)
-
-**之后**：项目只保留特定内容 (~10 个文件)
-
-```
-项目/.claude/
-├── MEMORY.md           # 项目记忆
-├── PROJECT_LOG.md      # 项目日志
-├── ANCHORS.md          # 项目锚点
-├── CLAUDE.md           # 项目配置 (继承+覆盖)
-├── thinking-routes/    # 思维轨迹
-└── rag/                # 项目 RAG 索引
+**Q: Codex can't read AGENTS.md**
+```bash
+# Regenerate from Claude config
+smc platform:convert claude codex
 ```
 
-### 收益
-
-| 指标 | 传统方式 | 分层继承 |
-|------|---------|---------|
-| 配置同步 | 手动复制到每个项目 | 自动（符号链接） |
-| 更新成本 | O(n) 项目数 | O(1) |
-| 配置漂移 | 容易发生 | 不可能 |
-| 存储空间 | 每项目 ~5MB | 每项目 ~50KB |
-
 ---
 
-## Documentation / 文档
+## Contributing
 
-- **[Development Guide / 开发指南](docs/DEVELOPMENT.md)** - Architecture, adding skills / 架构、添加技能
-- **[Marketplace Guide / 市场指南](docs/MARKETPLACE.md)** - Skill marketplace user guide / 技能市场用户指南
-- **[Q&A / 常见问题](Q&A.md)** - Core design concepts explained / 核心设计理念解释
+1. Fork the repository
+2. Create your feature branch (`git checkout -b feature/amazing`)
+3. Commit your changes (`git commit -m 'Add amazing feature'`)
+4. Push to the branch (`git push origin feature/amazing`)
+5. Open a Pull Request
 
 ---
 
-## Changelog / 更新日志
-
-### v1.2.0 (2026-01-17)
+## Documentation
 
-**English** | Comprehensive regression test suite.
+| Document | Description |
+|----------|-------------|
+| [Development Guide](docs/DEVELOPMENT.md) | Architecture and adding skills |
+| [Marketplace Guide](docs/MARKETPLACE.md) | Skill marketplace user guide |
+| [Q&A](Q&A.md) | Frequently asked questions |
 
-**中文** | 完整的回归测试套件。
+---
 
-- **Test Coverage / 测试覆盖**
-  - 219 new tests across 6 modules / 6 个模块新增 219 个测试
-  - errors.js: 100% statement coverage / 100% 语句覆盖率
-  - quality-gate.js: 95.59% statement coverage / 95.59% 语句覆盖率
-  - quality-rules.js: 86.17% statement coverage / 86.17% 语句覆盖率
-  - config-manager.js: 91.66% statement coverage / 91.66% 语句覆盖率
-  - config-validator.js: 77.41% statement coverage / 77.41% 语句覆盖率
-  - version-check.js: 55.07% statement coverage / 55.07% 语句覆盖率
+## Changelog
 
-### v1.1.1 (2026-01-16)
+### v1.6.0 (2026-01-24)
 
-**English** | Config validation and quality gate system.
+**Multi-Platform Support** - OpenAI Codex CLI compatibility
 
-**中文** | 配置验证和质量门禁系统。
+- **Platform Adapters** - Abstract layer for multi-CLI support
+- **Config Converter** - JSON ↔ TOML conversion
+- **Instruction Converter** - CLAUDE.md ↔ AGENTS.md
+- **New Commands**:
+  - `platform:detect` - Detect configured platforms
+  - `platform:list` - List supported platforms
+  - `platform:convert` - Convert between platforms
+  - `platform:sync` - Sync to all platforms
+- **Template Extensions**:
+  - `--codex` flag for Codex-only deployment
+  - `--all` flag for both platforms
 
-- **Config System / 配置系统**
-  - JSON Schema validation for all configs / 所有配置的 JSON Schema 验证
-  - Backup/rollback functionality / 备份回滚功能
-  - Environment variable expansion / 环境变量展开
-- **Quality Gate / 质量门禁**
-  - 8 built-in quality rules / 8 条内置质量规则
-  - Pre-commit/pre-push Git hooks / Git 钩子
-  - Configurable severity levels / 可配置严重级别
+### v1.5.2 (2026-01-23)
 
-### v1.1.0 (2026-01-15)
+- Architecture refactoring and code cleanup
+- Design standards system + 3 new skills
 
-**English** | Skill Marketplace with auto-sync.
+### v1.5.0 (2026-01-22)
 
-**中文** | 技能市场，支持自动同步。
+- Three-tier quality gate system
 
-- Marketplace system with 6 new commands / 市场系统，6 个新命令
-- Auto-sync via GitHub Actions / GitHub Actions 自动同步
-- 20+ built-in skills / 20+ 内置技能
+[Full Changelog](CHANGELOG.md)
 
 ---
 
-## License / 许可证
+## License
 
 MIT © [sumulige](https://github.com/sumulige)
 
 ---
 
-**Happy Coding with AI! / 祝编码愉快! 🚀**
+<p align="center">
+  <strong>Works with</strong><br>
+  <a href="https://claude.ai">Claude Code</a> •
+  <a href="https://openai.com/codex">Codex CLI</a>
+</p>

package/cli.js

@@ -9,7 +9,7 @@
  * - Built-in Claude Code project template
  * - ThinkingLens conversation tracking
  *
- * @version 1.5.2
+ * @version 1.6.0
  */
 
 const { runCommand } = require('./lib/commands');
@@ -37,8 +37,8 @@ const COMMANDS = {
     args: ''
   },
   template: {
-    help: 'Deploy Claude Code project template',
-    args: '[path] [--safe] [--force] [--help]'
+    help: 'Deploy AI CLI project template (Claude/Codex)',
+    args: '[path] [--codex] [--all] [--safe] [--force]'
   },
   kickoff: {
     help: 'Start project planning workflow (Manus-style)',
@@ -179,6 +179,22 @@ const COMMANDS = {
   audit: {
     help: 'Audit approved commands for security risks',
     args: '[--global] [--ci] [--report]'
+  },
+  'platform:detect': {
+    help: 'Detect AI CLI platforms in current project',
+    args: ''
+  },
+  'platform:list': {
+    help: 'List supported AI CLI platforms',
+    args: ''
+  },
+  'platform:convert': {
+    help: 'Convert config between platforms (claude ↔ codex)',
+    args: '<from> <to> [--dry-run]'
+  },
+  'platform:sync': {
+    help: 'Sync configuration to all platforms',
+    args: '[--platforms <list>]'
   }
 };
 

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "sumulige-claude",
-  "version": "1.5.2",
+  "version": "1.6.0",
   "description": "The Best Agent Harness for Claude Code - AI Agent Skills, Rules & Best Practices",
   "main": "cli.js",
   "bin": {
@@ -35,6 +35,9 @@
   "keywords": [
     "claude",
     "claude-code",
+    "codex",
+    "codex-cli",
+    "openai",
     "agent",
     "ai",
     "skills",
@@ -63,6 +66,7 @@
     "node": ">=16.0.0"
   },
   "dependencies": {
+    "@iarna/toml": "^2.2.5",
     "ajv": "^8.17.1",
     "ajv-formats": "^3.0.1",
     "patchright": "^1.57.0"

package/template/.codex/README.md

@@ -0,0 +1,69 @@
+# Codex CLI Configuration
+
+This directory contains configuration for [OpenAI Codex CLI](https://developers.openai.com/codex/).
+
+## Files
+
+| File | Purpose |
+|------|---------|
+| `config.toml` | Codex CLI settings (model, sandbox, approval policy) |
+| `../AGENTS.md` | Project instructions for the AI agent |
+
+## Quick Start
+
+```bash
+# Install Codex CLI
+npm install -g @openai/codex
+
+# Run Codex in your project
+codex
+
+# Or with a specific task
+codex "Fix the failing tests"
+```
+
+## Configuration Options
+
+### Sandbox Modes
+
+| Mode | Description |
+|------|-------------|
+| `read-only` | Can only read files, no modifications |
+| `workspace-write` | Can modify project files (recommended) |
+| `danger-full-access` | Full system access (use with caution) |
+
+### Approval Policies
+
+| Policy | Description |
+|--------|-------------|
+| `untrusted` | Require approval for all actions |
+| `on-failure` | Auto-approve, ask on errors (recommended) |
+| `on-request` | Auto-approve unless agent requests |
+| `never` | Never ask for approval |
+
+## Integration with Claude Code
+
+This project supports both Claude Code and Codex CLI:
+
+- Claude Code reads `.claude/CLAUDE.md` and `.claude/settings.json`
+- Codex CLI reads `AGENTS.md` and `.codex/config.toml`
+
+Both tools can use the same project rules in `.claude/rules/`.
+
+## Sync Configuration
+
+Use sumulige-claude to sync settings between platforms:
+
+```bash
+# Convert Claude config to Codex
+smc platform:convert claude codex
+
+# Sync all platforms
+smc platform:sync
+```
+
+## Documentation
+
+- [Codex CLI Documentation](https://developers.openai.com/codex/)
+- [AGENTS.md Guide](https://developers.openai.com/codex/guides/agents-md)
+- [Configuration Reference](https://developers.openai.com/codex/config-reference/)

package/template/.codex/config.toml

@@ -0,0 +1,56 @@
+# Codex CLI Configuration
+# Generated by sumulige-claude
+# Documentation: https://developers.openai.com/codex/config-reference/
+
+# Model Configuration
+model = "o3"
+model_provider = "openai"
+
+# Sandbox and Approval Settings
+# sandbox_mode options: read-only | workspace-write | danger-full-access
+sandbox_mode = "workspace-write"
+
+# approval_policy options: untrusted | on-failure | on-request | never
+approval_policy = "on-failure"
+
+# Project Settings
+[project]
+# Files to read as instructions (priority order)
+project_doc_fallback_filenames = ["AGENTS.md", "CLAUDE.md", "TEAM_GUIDE.md"]
+
+# Maximum size of instruction files (bytes)
+project_doc_max_bytes = 65536
+
+# Feature Flags
+[features]
+# Enable shell command execution
+shell_tool = true
+
+# Enable web search capability
+web_search_request = true
+
+# Enable unified PTY-backed exec (beta)
+# unified_exec = false
+
+# Enable shell environment snapshots (beta)
+# shell_snapshot = false
+
+# TUI Settings
+[tui]
+# Enable terminal animations
+animations = true
+
+# Show desktop notifications on completion
+notifications = true
+
+# MCP Server Configuration (uncomment to enable)
+# [mcp_servers.filesystem]
+# command = ["mcp-server-filesystem", "--root", "."]
+# enabled = true
+# enabled_tools = ["read_file", "write_file", "list_directory"]
+# env = {}
+
+# [mcp_servers.github]
+# command = ["mcp-server-github"]
+# enabled = true
+# enabled_tools = ["create_issue", "list_issues", "create_pr"]

package/template/AGENTS.md

@@ -0,0 +1,94 @@
+# Project Instructions
+
+> This file configures Codex CLI behavior for this project.
+> Generated by sumulige-claude. Edit to customize.
+> Documentation: https://developers.openai.com/codex/guides/agents-md
+
+---
+
+## Project Overview
+
+[Describe your project's purpose and goals here]
+
+---
+
+## Working Agreements
+
+- Run tests after modifications
+- Request approval before adding external dependencies
+- Document public API changes
+- Follow existing code patterns and conventions
+
+---
+
+## Code Standards
+
+### General Principles
+
+1. **Simplicity First** - Write clear, maintainable code
+2. **Test Coverage** - Maintain 80%+ test coverage
+3. **Documentation** - Document public APIs and complex logic
+
+### Style Guidelines
+
+Follow the project's established patterns:
+- See `.claude/rules/` for detailed guidelines (if present)
+- Run linting before committing
+
+---
+
+## Project Structure
+
+```
+project/
+├── src/           # Source code
+├── tests/         # Test files
+├── docs/          # Documentation
+├── .claude/       # Claude Code configuration
+└── .codex/        # Codex CLI configuration
+```
+
+---
+
+## Common Commands
+
+```bash
+# Development
+npm run dev        # Start development server
+npm test           # Run tests
+npm run build      # Build for production
+
+# Code Quality
+npm run lint       # Check code style
+npm run typecheck  # Check types
+```
+
+---
+
+## Codex-Specific Settings
+
+- **Sandbox Mode**: `workspace-write` - Can modify project files
+- **Approval Policy**: `on-failure` - Auto-approve unless errors occur
+- **Shell Tool**: Enabled for command execution
+- **Web Search**: Enabled for documentation lookup
+
+### Configuration
+
+Full configuration is in `.codex/config.toml`.
+
+### Fallback Files
+
+Codex also reads these instruction files if present:
+- `CLAUDE.md` - Claude Code instructions (compatible)
+- `TEAM_GUIDE.md` - Team guidelines
+
+---
+
+## Human Confirmation Required
+
+Request human confirmation before:
+- Architecture changes
+- Database schema modifications
+- Adding new external dependencies
+- Security-related changes
+- Deploying to production