@feng-h/pdca-skill 1.0.9 → 1.0.11

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (3) hide show
  1. package/CLAUDE.md +209 -93
  2. package/README.md +110 -82
  3. package/package.json +1 -1
package/CLAUDE.md CHANGED
@@ -2,107 +2,223 @@
2
2
 
3
3
  This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
4
4
 
5
- ## Project Overview
5
+ ---
6
6
 
7
- PDCA-with-AI is an **AI-driven workflow system** based on the PDCA (Plan-Do-Check-Act) cycle, not a traditional software codebase. The system integrates with Feishu (Lark) tools (Wiki, Bitable, Calendar, Task, Sheet) to provide structured problem-solving workflows for manufacturing and quality improvement scenarios.
7
+ ## 项目概述 / Project Overview
8
8
 
9
- **Key distinction:** This project consists primarily of markdown files that define AI agents, workflows, and templates. There is no TypeScript/JavaScript source code to build or test.
9
+ **PDCA-with-AI** 是一个基于 PDCA 循环的 AI 驱动型工作流系统,专为 **OpenClaw + 飞书(Feishu/Lark)** 深度集成设计。
10
10
 
11
- ## Architecture
11
+ **核心特点**:
12
+ - 非 traditional 软件代码库,主要由 markdown 文件定义 AI agents、workflows 和 templates
13
+ - 深度依赖飞书 API(Wiki、Bitable、Calendar、Task、Sheet)
14
+ - 遵循 [superpowers:writing-skills](https://github.com/obra/superpowers) 标准
12
15
 
13
- ### Entry Point
14
- - **`pdca/SKILL.md`** — The main skill definition. When the `pdca` skill is invoked, this file contains the primary logic for project evaluation and creation.
16
+ ---
15
17
 
16
- ### Core System Files (`_系统/`)
17
- - **`Agent/PDCA控制器.md`** — Central orchestrator managing the PDCA time flow, phase transitions, and rollback mechanisms
18
- - **`Agent/PlanAgent.md`**, **`DoAgent.md`**, **`CheckAgent.md`**, **`ActAgent.md`** — Phase-specific AI agents
19
- - **`规范/核心规范.md`** — Core system standards, including the principle that **Bitable is the single source of truth for all data**
20
- - **`规范/Validators/smart.md`** — SMART goal validation standards
21
- - **`规范/Validators/logic.md`** — Causal logic chain validation
18
+ ## 当前状态 / Current Status
22
19
 
23
- ### Reference Documentation (`pdca/references/`)
24
- - **`feishu-integration.md`** — Complete Feishu API integration patterns (Wiki, Bitable, Calendar, Task, Sheet)
25
- - **`cron-driving.md`** Automated inspection loop logic (The Loop: Scrape → Load → Verify → Message)
26
- - **`transition-checklist.md`** Phase transition requirements
27
- - **`exception-handling.md`** Exception handling patterns
20
+ | 项目 | 状态 |
21
+ |------|------|
22
+ | **GitHub** | https://github.com/Feng-H/PDCA-with-AI |
23
+ | **npm 包** | @feng-h/pdca-skill |
24
+ | **最新版本** | v1.0.9 |
25
+ | **安装方式** | `npx skills add Feng-H/PDCA-with-AI --agent skills` |
26
+
27
+ ---
28
+
29
+ ## 项目结构 / Project Structure
28
30
 
29
- ### Project Structure
30
31
  ```
31
32
  PDCA-with-AI/
32
- ├── pdca/ # Skill package (frontend interface layer)
33
- ├── SKILL.md # Main skill entry point
34
- ├── references/ # Detailed guides
35
- └── tests/ # Baseline tests
36
- ├── _系统/ # Core engine (backend specifications)
37
- ├── Agent/ # Agent definitions
38
- ├── 规范/ # Standards and validators
39
- └── 工具/ # Operational guides
40
- └── 项目/ # Active project instances
33
+ ├── SKILL.md # OpenClaw/Gemini 技能入口(主要文件)
34
+ ├── .claude/skills/ # Claude Code 参考文档(非执行)
35
+ ├── assets/ # 详细文档与测试
36
+ ├── references/ # 飞书 API 集成指南
37
+ ├── tests/ # 基线测试与质量验证
38
+ └── templates/ # 项目模板
39
+ ├── system/ # 核心引擎层
40
+ ├── Agent/ # Plan/Do/Check/Act Agent 提示词
41
+ │ ├── 规范/ # SMART 校验与因果逻辑模组
42
+ │ ├── 工具/ # Bitable 配置、工作流等
43
+ │ └── 模板/ # 系统模板
44
+ ├── bin/ # npx 可执行文件
45
+ ├── package.json # npm 包配置
46
+ └── README.md # 项目说明(中/英/日)
41
47
  ```
42
48
 
43
- ## Key Workflows
44
-
45
- ### Three Primary Commands
46
- The system is invoked through three main intents (defined in `_系统/项目管理系统.md`):
47
-
48
- | Command | Purpose | Output |
49
- |---------|---------|--------|
50
- | `new` | Initialize a new project | Creates Feishu Wiki + Bitable + Calendar resources |
51
- | `ongoing` | Manage active projects | Progress check + status update + alerts |
52
- | `achieve` | Search experience library | Best practice recommendations + template matching |
53
-
54
- ### Phase Transition Logic
55
- Phase transitions are **strictly sequential** and require:
56
- 1. Completion of all mandatory tasks in current phase
57
- 2. Verification against `_系统/工具/阶段转换检查表.md`
58
- 3. Explicit user confirmation (via Feishu interactive cards)
59
-
60
- ### The Loop (Automated Inspection)
61
- The PDCA Controller runs a 4-step inspection cycle:
62
- 1. **Scrape (Wiki)**: Fetch all project documents from Feishu Wiki
63
- 2. **Load (Logic)**: Load phase-specific "required conditions" and "deliverables checklist"
64
- 3. **Verify**: Compare Wiki content against requirements, identify gaps or logic conflicts
65
- 4. **Message (Card)**: Send Feishu interactive card (green/yellow/red based on severity)
66
-
67
- ## Quality Standards
68
-
69
- This project follows [superpowers:writing-skills](https://github.com/obra/superpowers) standards:
70
- - **RED-GREEN-REFACTOR**: All skills validated via baseline tests
71
- - **Rationalization Defense**: Explicit prohibitions against common agent shortcuts
72
- - **Red Flags**: Stop-and-restart signals when quality is at risk
73
-
74
- ### Critical Red Flags
75
- When you encounter these signals, **STOP immediately** and return to Plan phase:
76
- - Goals without specific numbers
77
- - "We'll adjust during execution"
78
- - "Quick fix is enough"
79
- - "Boss is rushing, let's start"
80
- - "Already spent X weeks" (sunk cost fallacy)
81
- - Accepting user-provided causes without Why/verification
82
-
83
- ### Rationalization Defenses
84
- Common excuses that must be rejected:
85
- - "Time is tight, need to launch quickly" → Fast launch without clear goals = accelerating in wrong direction
86
- - "No project manager, use simple approach" → Simple approach still needs SMART foundation
87
- - "User already identified the cause" → User provides symptoms, not root causes
88
- - "Plan took 3 weeks, too long" → Time is not a reason to skip quality checks
89
-
90
- ## Important Constraints
91
-
92
- 1. **Single Source of Truth**: Feishu Bitable is the only authoritative source for project status data
93
- 2. **User Decision Required**: All phase transitions and project closures require explicit user confirmation
94
- 3. **Immediate Alerting**: When inspection finds logic deviations or progress stagnation, immediately send Feishu interactive card
95
- 4. **No Shortcuts**: SMART validation and causal logic verification are mandatory, not optional
96
-
97
- ## Testing
98
-
99
- Baseline test scenarios are in `pdca/tests/baseline-test-scenarios.md`. These verify behavior WITHOUT the skill loaded to identify problems that the skill should solve.
100
-
101
- ## Feishu Integration Notes
102
-
103
- Key Feishu API patterns from `feishu-integration.md`:
104
- - **Bitable person fields**: Must be `[{id: "ou_xxx"}]`, not strings
105
- - **Bitable date fields**: Millisecond timestamps, not seconds
106
- - **Bitable hyperlink fields (type=15)**: Do NOT pass `property` parameter on create
107
- - **Calendar events**: Always pass `user_open_id` from SenderId so user can see the event
108
- - **No concurrent writes**: Bitable does not support concurrent writes, call serially with 0.5-1s delay
49
+ **路径变更说明**:
50
+ - 原 `pdca/` → 现 `assets/`
51
+ - `_系统/` → 现 `system/`
52
+
53
+ ---
54
+
55
+ ## 核心依赖 / Core Dependencies
56
+
57
+ ### 必需平台 / Required Platform
58
+
59
+ | 组件 | 说明 | 获取方式 |
60
+ |------|------|---------|
61
+ | **OpenClaw** | AI CLI 框架 | https://github.com/open-claw/open-claw |
62
+ | **飞书插件** | 提供 feishu-* API | OpenClaw 官方插件 |
63
+
64
+ ### 关键 API 列表
65
+
66
+ 本技能包依赖以下飞书 API(**仅 OpenClaw 飞书插件提供**):
67
+
68
+ | API | 功能 |
69
+ |-----|------|
70
+ | `feishu-create-doc` | 创建 Wiki 文档 |
71
+ | `feishu-update-doc` | 更新文档内容 |
72
+ | `feishu-fetch-doc` | 读取文档内容 |
73
+ | `feishu-bitable` | 多维表格 CRUD(27种字段类型) |
74
+ | `feishu-calendar` | 日程管理 |
75
+ | `feishu-task` | 任务管理 |
76
+ | `feishu-sheet` | 电子表格 |
77
+ | `feishu-im-read` | 消息读取 |
78
+
79
+ ---
80
+
81
+ ## 核心工作流 / Core Workflows
82
+
83
+ ### 三大命令
84
+
85
+ | 命令 | 功能 | 输出 |
86
+ |------|------|------|
87
+ | `new` | 启动新项目 | 创建 Wiki + Bitable 应用 + Calendar |
88
+ | `ongoing` | 管理活跃项目 | 进度检查 + 状态更新 + 预警 |
89
+ | `achieve` | 检索经验库 | 最佳实践推荐 + 模板匹配 |
90
+
91
+ ### 阶段转换逻辑
92
+
93
+ 阶段转换**严格按顺序**,需要:
94
+ 1. 当前阶段所有必选任务完成
95
+ 2. 对照 `system/工具/阶段转换检查表.md` 验证
96
+ 3. 用户通过飞书交互式卡片确认
97
+
98
+ ### The Loop(自治巡检)
99
+
100
+ PDCA 控制器运行的 4 步巡检循环:
101
+ 1. **Scrape (Wiki)**:获取所有项目文档
102
+ 2. **Load (Logic)**:加载阶段特定的"必需条件"和"交付物清单"
103
+ 3. **Verify**:对比 Wiki 内容与要求,识别差距或逻辑冲突
104
+ 4. **Message (Card)**:发送飞书交互式卡片(绿/黄/红基于严重程度)
105
+
106
+ ---
107
+
108
+ ## 质量标准 / Quality Standards
109
+
110
+ ### Red Flags - 立即停止
111
+
112
+ 遇到以下信号时,**立即停止**并返回 Plan 阶段:
113
+ - 目标没有具体数字
114
+ - "执行中再调整"
115
+ - "快速补救一下"
116
+ - "老板在催,先开始吧"
117
+ - "已经花了 X 周了"(沉没成本谬误)
118
+ - 直接接受用户提供的原因而不验证
119
+
120
+ ### Rationalization 防御
121
+
122
+ | 借口 | 现实 |
123
+ |------|------|
124
+ | "时间紧,需要快速上线" | 没有清晰目标的快速上线 = 朝错误方向加速 |
125
+ | "没有项目经理,用简单方式" | 简单方式仍需 SMART 基础 |
126
+ | "用户已经说了原因" | 用户说的是症状,不是根因 |
127
+
128
+ ---
129
+
130
+ ## 重要约束 / Important Constraints
131
+
132
+ 1. **Single Source of Truth**:飞书 Bitable 是项目状态的唯一权威数据源
133
+ 2. **User Decision Required**:所有阶段转换和项目结项需用户确认
134
+ 3. **Immediate Alerting**:巡检发现逻辑偏差或进度停滞,立即发送飞书交互式卡片
135
+ 4. **No Shortcuts**:SMART 校验和因果逻辑验证是强制的,不可跳过
136
+
137
+ ---
138
+
139
+ ## 飞书 API 注意事项 / Feishu API Notes
140
+
141
+ ### 关键坑位
142
+
143
+ | 字段类型 | 注意事项 |
144
+ |---------|---------|
145
+ | **人员字段** | 必须是 `[{id: "ou_xxx"}]`,不能是字符串 |
146
+ | **日期字段** | 毫秒时间戳,不是秒 |
147
+ | **超链接字段 (type=15)** | 创建时不要传 `property` 参数 |
148
+ | **日历事件** | 必须传 `user_open_id` 让用户能看到日程 |
149
+ | **并发写入** | Bitable 不支持并发写,串行调用 + 0.5-1s 延迟 |
150
+
151
+ ---
152
+
153
+ ## 测试 / Testing
154
+
155
+ 基线测试场景:`assets/tests/baseline-test-scenarios.md`
156
+
157
+ 这些验证**不加载技能时的行为**,以识别技能应该解决的问题。
158
+
159
+ ---
160
+
161
+ ## 使用后优化指南 / Optimization Guide
162
+
163
+ ### 当你发现问题时
164
+
165
+ 1. **问题分类**:
166
+ - **内容问题**(SKILL.md 描述不清、流程有误)
167
+ - **API 问题**(飞书 API 调用错误、参数不对)
168
+ - **逻辑问题**(阶段转换条件、验证规则)
169
+ - **格式问题**(文档结构、路径引用)
170
+
171
+ 2. **对应文件**:
172
+ | 问题类型 | 修改文件 |
173
+ |---------|---------|
174
+ | 内容问题 | `SKILL.md`, `system/Agent/*.md` |
175
+ | API 问题 | `assets/references/feishu-integration.md` |
176
+ | 逻辑问题 | `system/规范/Validators/*.md` |
177
+ | 格式问题 | `README.md`, `CLAUDE.md` |
178
+
179
+ 3. **修改后操作**:
180
+ ```bash
181
+ # 1. 提交到 GitHub
182
+ git add -A && git commit -m "fix: 描述修改内容"
183
+ git push origin main
184
+
185
+ # 2. 发布新版本
186
+ npm version patch # 或 minor/major
187
+ ```
188
+
189
+ 这会自动推送 + 发布到 npm(配置了 postversion hook)
190
+
191
+ ### 快速进入状态
192
+
193
+ 当你下次来优化时,需要检查:
194
+ 1. **npm 版本**:`npm view @feng-h/pdca-skill version`
195
+ 2. **文档结构**:`ls -la assets/ system/`
196
+ 3. **用户反馈**:具体问题描述
197
+
198
+ ---
199
+
200
+ ## 最近变更 / Recent Changes
201
+
202
+ | 版本 | 日期 | 变更 |
203
+ |------|------|------|
204
+ | v1.0.9 | 2026-04-06 | 添加 OpenClaw 对比表格,说明 Lark-CLI 可能性 |
205
+ | v1.0.8 | 2026-04-06 | 说明为什么需要 OpenClaw + 飞书 |
206
+ | v1.0.7 | 2026-04-06 | 添加中/英/日多语言 README |
207
+ | v1.0.6 | 2026-04-06 | 添加 Claude Code 技能支持 |
208
+ | v1.0.5 | 2026-04-06 | 添加 npx bin 可执行文件 |
209
+ | v1.0.0-v1.0.4 | 2026-04-06 | 初始版本,项目结构重构 |
210
+
211
+ ---
212
+
213
+ ## 发布流程 / Release Process
214
+
215
+ ```bash
216
+ # 一键发布(自动推送到 GitHub + npm)
217
+ npm run release:patch # 补丁版本 (1.0.9 → 1.0.10)
218
+ npm run release:minor # 次要版本 (1.0.9 → 1.1.0)
219
+ npm run release:major # 主要版本 (1.0.9 → 2.0.0)
220
+ ```
221
+
222
+ ---
223
+
224
+ **Designed for OpenClaw + Feishu** / **专为 OpenClaw + 飞书设计**
package/README.md CHANGED
@@ -1,9 +1,5 @@
1
1
  # PDCA with AI — OpenClaw + 飞书集成技能包
2
2
 
3
- [English](#english) | [中文](#中文) | [日本語](#日本語)
4
-
5
- ---
6
-
7
3
  <a href="https://www.npmjs.com/package/@feng-h/pdca-skill"><img src="https://img.shields.io/npm/v/@feng-h/pdca-skill" alt="npm version"></a>
8
4
  <a href="https://www.npmjs.com/package/@feng-h/pdca-skill"><img src="https://img.shields.io/npm/dm/@feng-h/pdca-skill" alt="npm downloads"></a>
9
5
  <a href="https://github.com/Feng-H/PDCA-with-AI/blob/main/LICENSE"><img src="https://img.shields.io/npm/l/@feng-h/pdca-skill" alt="License"></a>
@@ -11,19 +7,19 @@
11
7
 
12
8
  ---
13
9
 
14
- ## ⚠️ 重要说明 / Important Notice
10
+ ## [中文](#中文) | [English](#english) | [日本語](#日本語)
15
11
 
16
- **这是一个深度集成飞书 API 的专业技能包,必须配合 OpenClaw 飞书插件使用。**
12
+ ---
17
13
 
18
- This is a specialized skill package with **deep Feishu/Lark API integration**, designed exclusively for **OpenClaw with Feishu Plugin**.
14
+ <a name="中文"></a>
19
15
 
20
- **これは、Feishu/Lark API と深く統合された専門スキルパッケージであり、OpenClaw Feishu プラグインでの使用が必須です。**
16
+ ## 中文
21
17
 
22
- ---
18
+ ### ⚠️ 重要说明
23
19
 
24
- ## 为什么是 OpenClaw + 飞书?/ Why OpenClaw + Feishu?
20
+ **这是一个深度集成飞书 API 的专业技能包,必须配合 OpenClaw 飞书插件使用。**
25
21
 
26
- ### 中文
22
+ ### 为什么是 OpenClaw + 飞书?
27
23
 
28
24
  本技能包的核心功能依赖于以下**飞书专属 API**,这些 API 只有 OpenClaw 飞书插件提供:
29
25
 
@@ -39,17 +35,49 @@ This is a specialized skill package with **deep Feishu/Lark API integration**, d
39
35
 
40
36
  > *⚠️ 理论上可通过 Lark-CLI 等第三方工具实现,但未经测试。本技能包仅针对 OpenClaw 飞书插件进行验证。*
41
37
 
42
- **关键特性**:
43
- - **一个项目 = 一个文件夹 = 一个 Bitable 应用** — 飞书原生架构
44
- - **自治巡检 (The Loop)** AI 定期读取飞书文档、汇总 Bitable 数据
45
- - **交互式卡片预警** — 飞书消息卡片推送
46
- - **知识沉淀** — 自动归档到 Wiki 经验库
38
+ ### 🎯 系统简介
39
+
40
+ 基于 PDCA 循环的 AI 增强型问题解决工作流,专为 **OpenClaw + 飞书** 环境设计。
41
+
42
+ ### 核心亮点
47
43
 
48
- 这些功能都建立在飞书工具链的深度集成之上,无法在通用 AI 工具中实现。
44
+ - **自治巡检 (The Loop)**:AI 自动读取一线人员填写的云文档,汇总进展并同步至 Bitable 看板
45
+ - **质量校验 (Validator)**:在 Plan 阶段执行 **SMART 校验**与**因果逻辑链审查**
46
+ - **即时预警 (Real-time Cards)**:巡检发现逻辑偏差或进度停滞,AI 立即发送飞书交互式卡片
47
+ - **知识库 (Archiving)**:项目结项后,AI 自动提炼成功措施并归档
48
+ - **制造场景优化**:内置设备 OEE 提升、不良率降低等专业模板
49
+
50
+ ### 🚀 快速开始
51
+
52
+ **前提条件**:已安装 OpenClaw 并配置飞书插件
53
+
54
+ ```bash
55
+ # 通过 skills.sh 安装
56
+ npx skills add Feng-H/PDCA-with-AI --agent skills
57
+
58
+ # 或使用 Gemini CLI
59
+ gemini skills install https://github.com/Feng-H/PDCA-with-AI.git
60
+ ```
61
+
62
+ ### 常用指令
63
+
64
+ | 指令 | 说明 |
65
+ |------|------|
66
+ | `new` | 启动新项目 |
67
+ | `ongoing` | 查看活跃项目看板 |
68
+ | `achieve` | 检索历史成功案例 |
49
69
 
50
70
  ---
51
71
 
52
- ### English
72
+ <a name="english"></a>
73
+
74
+ ## English
75
+
76
+ ### ⚠️ Important Notice
77
+
78
+ **This is a specialized skill package with deep Feishu/Lark API integration, designed exclusively for OpenClaw with Feishu Plugin.**
79
+
80
+ ### Why OpenClaw + Feishu?
53
81
 
54
82
  This skill package relies on the following **Feishu-exclusive APIs**, only available via OpenClaw Feishu Plugin:
55
83
 
@@ -65,85 +93,79 @@ This skill package relies on the following **Feishu-exclusive APIs**, only avail
65
93
 
66
94
  > *⚠️ Theoretically achievable via third-party tools like Lark-CLI, but untested. This skill package is only validated for OpenClaw Feishu Plugin.*
67
95
 
68
- **Key Features**:
69
- - **One Project = One Folder = One Bitable App** — Native Feishu architecture
70
- - **Autonomous Inspection (The Loop)** — AI periodically reads Feishu docs, aggregates Bitable data
71
- - **Interactive Card Alerts** — Feishu message card push notifications
72
- - **Knowledge Archiving** — Auto-archive to Wiki experience library
96
+ ### 🎯 Overview
73
97
 
74
- These features are built on deep integration with Feishu toolchain and cannot be achieved in generic AI tools.
98
+ AI-enhanced problem-solving workflow based on the PDCA cycle, designed specifically for **OpenClaw + Feishu/Lark** environments.
75
99
 
76
- ---
100
+ ### Key Features
77
101
 
78
- ### 日本語
102
+ - **Autonomous Inspection (The Loop)**: AI automatically reads cloud documents, summarizes progress, and syncs to Bitable dashboards
103
+ - **Quality Validation**: SMART verification and causal logic chain review in Plan phase
104
+ - **Real-time Alerts**: AI sends Feishu interactive cards immediately when logic deviations or progress stagnation are detected
105
+ - **Knowledge Archiving**: AI extracts successful measures for future reuse
106
+ - **Manufacturing Optimization**: Built-in templates for OEE improvement, defect rate reduction, etc.
79
107
 
80
- このスキルパッケージは、以下の **Feishu 専用 API** に依存しており、OpenClaw Feishu プラグイン経由でのみ利用可能です:
108
+ ### 🚀 Quick Start
81
109
 
82
- | API 呼び出し | 機能 | OpenClaw | Claude Code | Gemini | Codex |
83
- |-------------|------|----------|-------------|--------|-------|
84
- | `feishu-create-doc` | Wiki/ナレッジベースにドキュメント作成 | ✅ | ⚠️* | ⚠️* | ⚠️* |
85
- | `feishu-update-doc` | ドキュメントの部分更新 | ✅ | ⚠️* | ⚠️* | ⚠️* |
86
- | `feishu-bitable` | Bitable CRUD(27種のフィールドタイプ) | ✅ | ⚠️* | ⚠️* | ⚠️* |
87
- | `feishu-calendar` | カレンダーとイベント管理 | ✅ | ⚠️* | ⚠️* | ⚠️* |
88
- | `feishu-task` | タスク作成、照会、更新 | ✅ | ⚠️* | ⚠️* | ⚠️* |
89
- | `feishu-sheet` | スプレッドシート操作 | ✅ | ⚠️* | ⚠️* | ⚠️* |
90
- | `feishu-im-read` | グループ/プライベートチャット履歴 | ✅ | ⚠️* | ⚠️* | ⚠️* |
110
+ **Prerequisite**: OpenClaw installed with Feishu plugin configured
91
111
 
92
- > *⚠️ 理論的には Lark-CLI などのサードパーティツールで実現可能ですが、未テストです。このスキルパッケージは OpenClaw Feishu プラグインでのみ検証されています。*
112
+ ```bash
113
+ # Install via skills.sh
114
+ npx skills add Feng-H/PDCA-with-AI --agent skills
115
+
116
+ # Or using Gemini CLI
117
+ gemini skills install https://github.com/Feng-H/PDCA-with-AI.git
118
+ ```
93
119
 
94
- **主な機能**:
95
- - **1プロジェクト = 1フォルダ = 1 Bitable アプリ** — Feishu ネイティブアーキテクチャ
96
- - **自律検査(The Loop)** — AI が定期的に Feishu ドキュメントを読み取り、Bitable データを集約
97
- - **インタラクティブカードアラート** — Feishu メッセージカードプッシュ通知
98
- - **知識アーカイブ** — Wiki 経験ライブラリへ自動アーカイブ
120
+ ### Common Commands
99
121
 
100
- これらの機能は Feishu ツールチェーンとの深い統合に基づいており、汎用 AI ツールでは実現できません。
122
+ | Command | Description |
123
+ |---------|-------------|
124
+ | `new` | Start a new project |
125
+ | `ongoing` | View active project dashboard |
126
+ | `achieve` | Search historical success cases |
101
127
 
102
128
  ---
103
129
 
104
- ## 🚀 快速开始 / Quick Start
130
+ <a name="日本語"></a>
105
131
 
106
- ### 中文
132
+ ## 日本語
107
133
 
108
- **前提条件**:已安装 OpenClaw 并配置飞书插件
134
+ ### ⚠️ 重要なお知らせ
109
135
 
110
- ```bash
111
- # 通过 skills.sh 安装
112
- npx skills add Feng-H/PDCA-with-AI --agent skills
113
-
114
- # 或使用 Gemini CLI
115
- gemini skills install https://github.com/Feng-H/PDCA-with-AI.git
116
- ```
136
+ **これは、Feishu/Lark API と深く統合された専門スキルパッケージであり、OpenClaw Feishu プラグインでの使用が必須です。**
117
137
 
118
- **核心命令**:
119
- - `new` — 启动新 PDCA 项目(自动创建 Wiki + Bitable + Calendar)
120
- - `ongoing` — 查看活跃项目看板
121
- - `achieve` — 检索历史成功案例
138
+ ### なぜ OpenClaw + Feishu なのか?
122
139
 
123
- ---
140
+ このスキルパッケージは、以下の **Feishu 専用 API** に依存しており、OpenClaw Feishu プラグイン経由でのみ利用可能です:
124
141
 
125
- ### English
142
+ | API 呼び出し | 機能 | OpenClaw | Claude Code | Gemini | Codex |
143
+ |-------------|------|----------|-------------|--------|-------|
144
+ | `feishu-create-doc` | Wiki/ナレッジベースにドキュメント作成 | ✅ | ⚠️* | ⚠️* | ⚠️* |
145
+ | `feishu-update-doc` | ドキュメントの部分更新 | ✅ | ⚠️* | ⚠️* | ⚠️* |
146
+ | `feishu-bitable` | Bitable CRUD(27種のフィールドタイプ) | ✅ | ⚠️* | ⚠️* | ⚠️* |
147
+ | `feishu-calendar` | カレンダーとイベント管理 | ✅ | ⚠️* | ⚠️* | ⚠️* |
148
+ | `feishu-task` | タスク作成、照会、更新 | ✅ | ⚠️* | ⚠️* | ⚠️* |
149
+ | `feishu-sheet` | スプレッドシート操作 | ✅ | ⚠️* | ⚠️* | ⚠️* |
150
+ | `feishu-im-read` | グループ/プライベートチャット履歴 | ✅ | ⚠️* | ⚠️* | ⚠️* |
126
151
 
127
- **Prerequisite**: OpenClaw installed with Feishu plugin configured
152
+ > *⚠️ 理論的には Lark-CLI などのサードパーティツールで実現可能ですが、未テストです。このスキルパッケージは OpenClaw Feishu プラグインでのみ検証されています。*
128
153
 
129
- ```bash
130
- # Install via skills.sh
131
- npx skills add Feng-H/PDCA-with-AI --agent skills
154
+ ### 🎯 概要
132
155
 
133
- # Or using Gemini CLI
134
- gemini skills install https://github.com/Feng-H/PDCA-with-AI.git
135
- ```
156
+ PDCAサイクルに基づく AI 強化型問題解決ワークフロー、**OpenClaw + Feishu/Lark** 環境のために特別に設計されました。
136
157
 
137
- **Core Commands**:
138
- - `new` — Start a new PDCA project (auto-creates Wiki + Bitable + Calendar)
139
- - `ongoing` — View active project dashboard
140
- - `achieve` — Search historical success cases
158
+ ### 主な機能
141
159
 
142
- ---
160
+ - **自律検査(The Loop)**:AI が自動的にクラウドドキュメントを読み取り、進捗を要約して Bitable ダッシュボードに同期
161
+ - **品質検証**:Plan フェーズでの SMART 検証と因果ロジックチェーンレビュー
162
+ - **リアルタイムアラート**:論理の偏差や進捗の停滞を検出すると、AI が即座に Feishu インタラクティブカードを送信
163
+ - **知識アーカイブ**:AI が成功した施策を抽出して将来のために保存
164
+ - **製造業向け最適化**:OEE 向上、不良率低減などの専門テンプレートを内蔵
143
165
 
144
- ### 日本語
166
+ ### 🚀 クイックスタート
145
167
 
146
- **前提条件**: OpenClaw がインストールされ、Feishu プラグインが設定されていること
168
+ **前提条件**:OpenClaw がインストールされ、Feishu プラグインが設定されていること
147
169
 
148
170
  ```bash
149
171
  # skills.sh 経由でインストール
@@ -153,32 +175,38 @@ npx skills add Feng-H/PDCA-with-AI --agent skills
153
175
  gemini skills install https://github.com/Feng-H/PDCA-with-AI.git
154
176
  ```
155
177
 
156
- **主要コマンド**:
157
- - `new` — 新規 PDCA プロジェクトを開始(Wiki + Bitable + Calendar を自動作成)
158
- - `ongoing` アクティブなプロジェクトダッシュボードを表示
159
- - `achieve` — 過去の成功事例を検索
178
+ ### 一般コマンド
179
+
180
+ | コマンド | 説明 |
181
+ |---------|------|
182
+ | `new` | 新規プロジェクトを開始 |
183
+ | `ongoing` | アクティブなプロジェクトダッシュボードを表示 |
184
+ | `achieve` | 過去の成功事例を検索 |
160
185
 
161
186
  ---
162
187
 
163
- ## 📁 项目结构 / Project Structure
188
+ ## 📁 Project Structure
164
189
 
165
190
  ```
166
191
  PDCA-with-AI/
167
192
  ├── SKILL.md # OpenClaw/Gemini skill entry point
168
- ├── .claude/skills/ # Reference documentation (not for execution)
193
+ ├── .claude/skills/ # Reference documentation
169
194
  ├── assets/ # Documentation and tests
170
195
  │ ├── references/ # Feishu API integration guides
171
196
  │ ├── tests/ # Quality validation
172
197
  │ └── templates/ # Project templates
173
198
  ├── system/ # Core engine layer
174
199
  │ ├── Agent/ # Plan/Do/Check/Act Agent prompts
175
- └── 规范/ # SMART validation modules
200
+ ├── 规范/ # SMART validation modules
201
+ │ ├── 工具/ # Bitable config, workflows
202
+ │ └── 模板/ # System templates
203
+ ├── bin/ # npx executable
176
204
  └── README.md
177
205
  ```
178
206
 
179
- ## 📚 文档 / Documentation
207
+ ## 📚 Documentation
180
208
 
181
- - [飞书 API 集成指南](assets/references/feishu-integration.md) - Complete Feishu API Integration Guide
209
+ - [飞书 API 集成指南](assets/references/feishu-integration.md) - Complete Feishu/Lark API Integration Guide
182
210
  - [OpenClaw 项目](https://github.com/open-claw/open-claw) - OpenClaw Plugin
183
211
  - [质量标准](assets/tests/) - Quality Standards & Testing
184
212
 
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "@feng-h/pdca-skill",
3
- "version": "1.0.9",
3
+ "version": "1.0.11",
4
4
  "description": "PDCA project management system for Feishu/Lark integration with AI-driven workflow",
5
5
  "keywords": [
6
6
  "skills",