@chongyan/autospec 1.0.1 → 1.0.2

package/knowledge/{principles/evolution.md → 01-principles/10-evolution-mechanism.md}

@@ -2,8 +2,8 @@
 
 > 核心思想：**让框架从每次执行中学习，但进化本身必须受约束、可回滚、可审计。**
 >
-> **版本**: v2.0
-> **最后更新**: 2026-03-26
+> **版本**: v2.1
+> **最后更新**: 2026-03-27
 > **基于**: FinSDD 自进化体系、Claude Marketplace 插件进化机制
 
 ## 三层自进化体系
@@ -15,7 +15,7 @@ AutoSpec 的自进化在三个层面分层发生，形成从项目实践到框
 │  L1: Self-Evolve (框架级)                                        │
 │  /self-evolve                                                    │
 │  目标: AutoSpec 框架自身的迭代升级                                  │
-│  位置: .claude/skills/self-evolve/                               │
+│  位置: .claude/commands/self-evolve.md                          │
 │  产出: 框架技能、方法论、设计原则                                  │
 └─────────────────────────────────────────────────────────────────┘
                               ↓ 提炼
@@ -23,7 +23,7 @@ AutoSpec 的自进化在三个层面分层发生，形成从项目实践到框
 │  L2: Field-Evolve (领域级)                                        │
 │  /autospec:field-evolve                                          │
 │  目标: 框架从实战项目中学习和提炼能力                               │
-│  位置: plugins/commands/autospec:field-evolve.md                            │
+│  位置: plugins/commands/field-evolve.md                            │
 │  产出: 领域技能、最佳实践、benchmarks                              │
 └─────────────────────────────────────────────────────────────────┘
                               ↓ 应用
@@ -31,7 +31,7 @@ AutoSpec 的自进化在三个层面分层发生，形成从项目实践到框
 │  L3: Project-Evolve (项目级)                                      │
 │  /autospec:project-evolve                                        │
 │  目标: 实战项目自身的代码、配置、文档优化                          │
-│  位置: plugins/commands/autospec:project-evolve.md                        │
+│  位置: plugins/commands/project-evolve.md                        │
 │  产出: 代码质量提升、测试补全、文档完善                            │
 └─────────────────────────────────────────────────────────────────┘
 ```
@@ -129,7 +129,7 @@ L3 Project-Evolve（升级后的框架用于新项目）
 | -------------------------------------------------- | ------------- |
 | `constitution.md`                                  | 安全红线，是整个系统的底座 |
 | `evolution.md`（本文件）                                | 进化规则本身，防止自我篡改 |
-| `pipeline/overview.md` 的启动门禁和双层验证定义                | 流程骨架，牵一发动全身   |
+| `knowledge/02-process/00-overview.md` 的启动门禁和双层验证定义 | 流程骨架，牵一发动全身   |
 | `methodology/design-philosophy.md` 的第一性原则（DP1-DP3） | 根基原则          |
 | 评判标准的判定逻辑（审类skill的"必须满足"条款）                        | 防止降低标准来通过验证   |
 
@@ -145,8 +145,8 @@ L3 Project-Evolve（升级后的框架用于新项目）
 
 | 内容                                                    | 进化方式                   |
 | ----------------------------------------------------- | ---------------------- |
-| `pipeline/01-requirement.md` \~ `06-delivery.md` 阶段定义 | 基于practice-log分析提出优化建议 |
-| `skills/stages/` 通用技能                                 | 基于执行效果提出改进             |
+| `knowledge/02-process/01-requirement.md` \~ `06-delivery.md` 阶段定义 | 基于practice-log分析提出优化建议 |
+| `skills/` 通用技能（requirement-stage, design-stage等）       | 基于执行效果提出改进             |
 | `roles/{role-name}.md` 角色画像                           | 基于实际项目反馈调整关注点          |
 | `methodology/principles.md`                           | 基于实践验证新增/修订原则          |
 | `methodology/checklists/`                             | 基于审查结果补充检查项            |
@@ -181,7 +181,7 @@ L3 Project-Evolve（升级后的框架用于新项目）
 
 **环境衔接层知识，按类型分属不同进化分区。**
 
-详见 `knowledge/environment/overview.md`。
+详见 `knowledge/06-environment/README.md`。
 
 | 内容 | 进化分区 | 进化方式 |
 |------|---------|---------|
@@ -210,14 +210,14 @@ L3 Project-Evolve（升级后的框架用于新项目）
    - 配置变更需经过验证流程
 
 4. **测试知识进化**：
-   - 触发时机：每次交付完成后，skill-distiller 扫描测试执行记录
+   - 触发时机：每次交付完成后，field-evolve-distiller 扫描测试执行记录
    - 学习信号：
      - 新的测试框架或工具使用模式（出现3次以上）
      - 更高效的 Mock 策略或测试数据准备方法
      - 测试失败模式重复出现（同类失败3次以上）
      - 测试代码风格趋于稳定（可提炼为模板）
    - 进化流程：
-     - skill-distiller 识别测试模式 → 自动生成测试知识提案
+     - field-evolve-distiller 识别测试模式 → 自动生成测试知识提案
      - 提案存放于 `.autospec/evolution/proposals/` 待审批
      - 人工审批后写入 `environment/integration/testing/`
    - 知识格式：推荐使用自由的 Markdown 格式，而非结构化 YAML
@@ -238,11 +238,64 @@ L3 Project-Evolve（升级后的框架用于新项目）
     │              │            │            │            │           └─ 拒绝 = 回滚
     │              │            │            │            └─ skill-validator + 基准对比
     │              │            │            └─ 区分冻结/受控/自由区
-    │              │            └─ skill-distiller + methodology-extractor
+    │              │            └─ field-evolve-distiller + methodology-extractor
     │              └─ practice-log + 执行指标
     └─ pipeline-runner 驱动
 ```
 
+## 三层自进化隔离演进
+
+### 隔离原则
+
+| 原则 | 说明 |
+|------|------|
+| **数据隔离** | 每个进化的practice-log独立存储 |
+| **执行隔离** | 每个进化在独立环境执行 |
+| **回滚隔离** | 每个进化独立回滚，不影响其他 |
+| **评测隔离** | 每个进化独立评测，不共享指标 |
+
+### 隔离实现架构
+
+```
+L1 Self-Evolve（框架级）→ /self-evolve
+├── 独立环境：.claude/self-evolve/
+├── 独立评测：.claude/benchmarks/self-evolve/
+├── 独立回滚：git branch self-evolve-{timestamp}
+├── 独立日志：.autospec/logs/self-evolve/
+└── 触发条件：人工触发 + 完整评测通过
+
+L2 Field-Evolve（领域级）→ /autospec:field-evolve
+├── 独立环境：.autospec/evolution/field/
+├── 独立评测：.autospec/benchmarks/field/
+├── 独立回滚：.autospec/evolution/rollback/field/
+├── 独立日志：.autospec/logs/field-evolve/
+└── 触发条件：5条practice-log累积 + 评测通过
+
+L3 Project-Evolve（项目级）→ /autospec:project-evolve
+├── 独立环境：项目目录/.autospec/
+├── 独立评测：项目目录/.autospec/benchmarks/
+├── 独立回滚：项目目录/.autospec/evolution/rollback/
+├── 独立日志：项目目录/.autospec/logs/
+└── 触发条件：每次交付后自动触发
+```
+
+### 进化前评测（强制）
+
+| 评测类型 | 内容 | 作用 |
+|---------|------|------|
+| **语法评测** | 验证修改内容语法正确 | 防止格式错误 |
+| **语义评测** | 验证修改内容语义合理 | 防止逻辑错误 |
+| **一致性评测** | 验证与冻结区/受控区无冲突 | 防止区域冲突 |
+| **安全评测** | 验证无危险操作 | 防止安全风险 |
+
+### 进化后评测（强制）
+
+| 评测类型 | 内容 | 作用 |
+|---------|------|------|
+| **回归评测** | 验证现有功能未被破坏 | 防止退化 |
+| **效果评测** | 验证改进目标达成 | 确认有效 |
+| **稳定性评测** | 验证多次执行稳定 | 防止震荡 |
+
 ### 基于评测驱动的自进化
 
 AutoSpec 通过 **Benchmarks 评测** 驱动进化，形成"评测 → 分析 → 改进 → 验证"的闭环：
@@ -458,9 +511,10 @@ Claude Code 原生机制          AutoSpec 用途
 | -------------------------- | ----------------------- | ----------------------------------- |
 | `frozen-zone-guard`        | PreToolUse (Write/Edit) | 阻止对冻结区文件的自动修改                       |
 | `scope-sentinel`           | PreToolUse (Write/Edit) | 编码阶段偏航检测（三级递增提醒）                    |
-| `layer1-validator`         | PreToolUse/PostToolUse  | Layer 1 强制检查 + 结果捕获 + 质量追踪           |
+| `layer1-validator`         | PreToolUse/PostToolUse  | Layer 1 强制检查 + 结果捕获 + 质量追踪 + **自主修复循环** |
 | `execution-tracker`        | PostToolUse (全部)        | 追踪执行日志（skill 加载、agent 调用、产物生成）       |
 | `environment-manager`      | Stop                    | 环境知识验证 + 未注册中间件检测                   |
+| `auto-fix-loop`            | PostToolUse (Layer 1 失败) | 自主修复循环（依赖/配置/代码/资源错误自动修复）            |
 
 Hooks 是**确定性的**（shell脚本，非AI判断），对应 DP1 第一层验证理念。
 企业可在此基础上追加自定义 hooks（合规检查、审计日志等）。
@@ -476,7 +530,7 @@ Hooks 是**确定性的**（shell脚本，非AI判断），对应 DP1 第一层
 | `implementation-stage` | 编码实现场景                            | 加载实现流程 + skill + Brownfield适配 + 角色关注点                    |
 | `testing-stage`      | 测试场景                              | 加载测试流程 + skill + 角色特有测试关注点                               |
 | `delivery-stage`   | 交付部署场景                            | 加载交付流程 + 安全检查 + 自动触发practice-log和进化信号                    |
-| `evolution-process`    | 框架进化/流程完成后                        | 加载进化流程（强制执行 distiller + extractor + evolver + validator） |
+| `field-evolve-*`       | 框架进化/流程完成后                        | 加载进化流程（scanner → executor → analyzer → fixer → verifier → learner → distiller） |
 | `pending-dashboard`   | 新会话启动时                            | 扫描待处理提案/practice-log/中断流程，主动提醒用户                         |
 
 Skills 是**被动激活的**——Claude 根据 description 自动判断是否加载。
@@ -551,3 +605,95 @@ Skills 是**被动激活的**——Claude 根据 description 自动判断是否
 }
 ```
 
+## 自主修复机制
+
+AutoSpec 框架具备**自主修复能力**，在遇到错误时先尝试自动修复，而不是立即标记为 BLOCKED。
+
+### 修复循环流程
+
+```
+检测障碍 → 分析根因 → 决策策略 → 执行修复 → 验证结果
+    ↑                                      │
+    └──────────── 最多 3 次循环 ─────────────┘
+```
+
+### 错误分类与修复策略
+
+| 错误类型 | 自主修复策略 | 示例 |
+|---------|------------|------|
+| **依赖类** | 自动添加/安装依赖 | 缺少 npm/Maven 包、工具缺失 |
+| **配置类** | 自动修正配置 | 配置文件错误、环境变量缺失 |
+| **代码类** | 自动修正代码 | 语法错误、类型错误、导入缺失 |
+| **资源类** | 自动创建/下载 | 文件缺失、目录不存在 |
+| **验证类** | 自动重试/修正 | 测试失败、编译错误 |
+| **逻辑类** | 需人工确认 | 业务逻辑错误、需求歧义 |
+
+### 修复边界
+
+| 区域 | 内容 | 修复方式 |
+|------|------|---------|
+| **冻结区** | 宪法文件、核心原则 | 不可自主修改，直接 BLOCKED |
+| **受控区** | 框架核心机制、接口契约 | 需人工批准，生成修复提案 |
+| **自由区** | 依赖添加、配置修正、语法错误 | 可自主修复，修复后验证 |
+
+### 实现位置
+
+- **核心逻辑**: `plugins/hooks/lib/auto-fix-loop.js`
+- **集成点**: `plugins/hooks/layer1-validator.js`（Layer 1 失败时触发）
+- **实践日志**: `.autospec/practice-log/`（记录每次修复详情）
+
+### 集成方式
+
+自主修复能力已集成到 `layer1-validator` hook 中：
+
+```
+Layer 1 验证失败
+    ↓
+analyzeErrorType() 分析错误类型
+    ↓
+getRepairZone() 判断修复区域
+    ↓
+自由区？→ autoFixLoop() 执行修复循环
+    ↓
+验证修复结果
+    ↓
+成功 → 继续执行 | 失败 → 重试（最多 3 次）→ 升级人工
+```
+
+### 实践日志记录
+
+每次自主修复都会记录到 practice-log：
+
+```markdown
+## [日期] 自主修复记录
+
+**问题**: [问题描述]
+**根因**: [错误类型]
+**修复方案**: [具体修复步骤]
+**验证结果**: [pass/fail]
+**分类**: [依赖/配置/代码/资源/验证]
+**区域**: [frozen/controlled/free]
+**重试次数**: [0-3]
+```
+
+### 安全护栏
+
+1. **最大重试 3 次** - 同一问题修复 3 次后必须升级人工
+2. **冻结区不可碰** - 宪法文件、核心原则不可自主修改
+3. **必须验证** - 修复后必须重新验证，不可跳过
+4. **完整记录** - 所有修复必须记录到 practice-log，用于进化分析
+
+### 与自进化的关系
+
+自主修复机制是 L3 Project-Evolve（项目级）的核心能力：
+
+```
+项目执行 → 遇到错误 → 自主修复 → 成功 → practice-log
+                                    ↓
+              同类问题 3 次重复 → field-evolve 提炼 → 通用技能
+                                    ↓
+                                    框架能力增强
+```
+
+自主修复的成功案例会通过 `field-evolve` 提炼为通用技能，增强框架的整体修复能力。
+

package/knowledge/01-principles/README.en.md

@@ -0,0 +1,385 @@
+# AutoSpec Principles System
+
+> **Purpose**: The framework's "principle operating system" — all execution decisions must trace back to principles in this document
+>
+> **Evolution Zone**: 🔴 Frozen Zone (Constitution, Evolution Rules, Core Principles, Design Philosophy)
+
+---
+
+## Principles System Overview
+
+```
+┌─────────────────────────────────────────────────────────────┐
+│  L1: First Principles — Inviolable "Laws of Physics"         │
+│  ─────────────────────────────────────────────────────────  │
+│  Verification-First | Minimal-Viable | Progressive-Autonomy  │
+│  → Located at: 01-first-principles.md                         │
+└─────────────────────────────────────────────────────────────┘
+                              │ Deduction
+┌─────────────────────────────────────────────────────────────┐
+│  L2: Strategic Principles — Directional Guidance             │
+│  ─────────────────────────────────────────────────────────  │
+│  Two-Layer Validation | Do-Review Separation | Harness Eng.  │
+│  → Located at: 02-strategic-principles.md                     │
+└─────────────────────────────────────────────────────────────┘
+                              │ Expansion
+┌─────────────────────────────────────────────────────────────┐
+│  L3: Tactical Principles — Execution Strategy Guidance       │
+│  ─────────────────────────────────────────────────────────  │
+│  Task Breakdown | Context Management | Precision Rollback    │
+│  → Located at: 03-tactical-principles.md                      │
+└─────────────────────────────────────────────────────────────┘
+                              │ Implementation
+┌─────────────────────────────────────────────────────────────┐
+│  L4: Operational Principles — Concrete Behavior Standards    │
+│  ─────────────────────────────────────────────────────────  │
+│  Coding Style | Auto-Fix | Constitution Gate | Checklists   │
+│  → Located at: 04-operational-principles.md                   │
+└─────────────────────────────────────────────────────────────┘
+```
+
+---
+
+## Directory Structure
+
+```
+principles/
+├── README.md                         # This file (Principles Navigation)
+├── 00-principles-hierarchy.md        # Principles Hierarchy (L1→L2→L3→L4 Traceability)
+├── 01-first-principles.md            # L1 First Principles (P1-P3 Triangular Prism)
+├── 02-strategic-principles.md        # L2 Strategic Principles (S1-S4)
+├── 03-tactical-principles.md         # L3 Tactical Principles (T1-T6)
+├── 04-operational-principles.md      # L4 Operational Principles (O1-O4)
+├── 05-domain-principles.md           # Domain Practice Principles (DDD/System/Requirement/Coding/CR/KM)
+├── 06-methodology-principles.md      # Methodology Principles (P1-P44)
+├── 07-cognitive-principles.md        # Cognitive Enhancement Principles (CP1-CP10)
+├── 08-auto-fix-principles.md         # Auto-Fix Principles (Extracted from Practice)
+├── 09-constitution.md                # Project Constitution (15 Red Lines, Stop on Violation)
+└── 10-evolution-mechanism.md         # Three-Layer Evolution Mechanism
+```
+
+---
+
+## Core Principle Documents
+
+### 00-principles-hierarchy.md — Principles Hierarchy
+
+**Content**: Complete L1→L2→L3→L4 traceability chain, principle application matrix, quick reference table
+
+**Core Chapters**:
+- Principles System Overview (L1→L2→L3→L4)
+- Principle Traceability Chain
+- Principle Application Matrix (by Stage / by Error Type)
+- Principle Quick Reference Table
+- Evolution Zone Description
+
+**Location**: [00-principles-hierarchy.md](00-principles-hierarchy.md)
+
+### 01-first-principles.md — L1 First Principles
+
+**Content**: AutoSpec's "Laws of Physics" — all upper-level principles are deduced from here
+
+**Core Chapters**:
+- P1 Verification-First
+- P2 Minimal-Viable Process
+- P3 Progressive Autonomy
+- Core Design Patterns (Two-Step Generation, Progressive Disclosure, Do-Review Separation, Harness Engineering)
+
+**Location**: [01-first-principles.md](01-first-principles.md)
+
+### 02-strategic-principles.md — L2 Strategic Principles
+
+**Content**: Directional guidance principles deduced from First Principles
+
+**Core Chapters**:
+- S1 Two-Layer Validation
+- S2 Do-Review Separation (Evaluator-Optimizer Separation)
+- S3 Harness Engineering Trilogy
+- S4 Informed Guessing
+
+**Location**: [02-strategic-principles.md](02-strategic-principles.md)
+
+### 03-tactical-principles.md — L3 Tactical Principles
+
+**Content**: Execution strategy guidance principles expanded from Strategic Principles
+
+**Core Chapters**:
+- T1 Task Breakdown Quality (ATOMIC Principles)
+- T2 Context Precision Management
+- T3 Precision Rollback
+- T4 Anti-Pattern Inoculation
+- T5 Taming Non-Determinism
+- T6 Prioritize Reusing Project Capabilities
+
+**Location**: [03-tactical-principles.md](03-tactical-principles.md)
+
+### 04-operational-principles.md — L4 Operational Principles
+
+**Content**: Concrete implementation of Tactical Principles, including specific behavior standards and checklists
+
+**Core Chapters**:
+- O1 Coding Style
+- O2 Auto-Fix Mechanism (Auto-Fix Loop)
+- O3 Constitution Gate
+- O4 Checklists
+
+**Location**: [04-operational-principles.md](04-operational-principles.md)
+
+### 05-domain-principles.md — Domain Practice Principles
+
+**Content**: General principles extracted from domain practices, guiding execution decisions in specific domains
+
+**Core Chapters**:
+- DDP1-4: DDD Tactical Design Principles
+- SDP1-3: System Design Principles
+- REP1-4: Requirement Engineering Principles
+- CEP1-6: Coding Practice Principles
+- CRP1-5: Code Review Principles
+- KMP1-4: Knowledge Management Principles
+
+**Location**: [05-domain-principles.md](05-domain-principles.md)
+
+### 06-methodology-principles.md — Methodology Principles
+
+**Content**: General principles extracted from knowledge and practice (P1-P44)
+
+**Core Chapters**:
+- Thinking Principles (P1-P6, P30-P31)
+- Design Principles (P7-P13, P32-P34)
+- Product & Requirement Principles (P14-P18)
+- Engineering & Quality Principles (P19-P21)
+- Governance & Evolution Principles (P22-P29)
+- Architecture Implementation Principles (P35-P44)
+
+**Location**: [06-methodology-principles.md](06-methodology-principles.md)
+
+### 07-cognitive-principles.md — Cognitive Enhancement Principles
+
+**Content**: Enhancing AI's cognitive intelligence level in the automated development process (CP1-CP10)
+
+**Core Chapters**:
+- Collaboration Stance (CP1 Partner Mode, CP2 Context Matching)
+- Cognitive Discipline (CP3 Ask Before Act, CP4 Complete Logic Chain, CP5 Confidence Labeling)
+- Error & Learning (CP6 Error as Signal, CP7 Pattern Recognition)
+- Deviation Detection (CP8 Three-Level Deviation Alerts)
+- Output Quality (CP9 Just Enough, CP10 Self-Check)
+
+**Location**: [07-cognitive-principles.md](07-cognitive-principles.md)
+
+### 08-auto-fix-principles.md — Auto-Fix Principles
+
+**Content**: Principles extracted from auto-fix mechanism practice
+
+**Core Chapters**:
+- Core Insights (3 items)
+- Principles System (AFP1-AFP6)
+- Principle Application Matrix
+- Principle Checklist
+
+**Location**: [08-auto-fix-principles.md](08-auto-fix-principles.md)
+
+### 09-constitution.md — Project Constitution
+
+**Content**: Red line checklist — violating any item immediately stops the entire process
+
+**Core Chapters**:
+- Safety Red Lines (#1-#3)
+- Process Red Lines (#4-#9)
+- Decision Red Lines (#10-#11)
+- Stop Conditions
+- Auto-Fix Mechanism
+
+**Location**: [09-constitution.md](09-constitution.md)
+
+### 10-evolution-mechanism.md — Self-Evolution Mechanism
+
+**Content**: Three-layer self-evolution system, complete closed loop from project practice to framework upgrade
+
+**Core Chapters**:
+- L1: Self-Evolve (Framework Self-Evolution)
+- L2: Field-Evolve (Domain Evolution)
+- L3: Project-Evolve (Project Self-Evolution)
+- Evolution Zones (Frozen / Controlled / Free)
+- Safety Guardrails
+
+**Location**: [10-evolution-mechanism.md](10-evolution-mechanism.md)
+
+---
+
+## Quick Navigation
+
+### Find Principles by Scenario
+
+| Scenario | Applicable Principles | Location |
+|----------|----------------------|----------|
+| **Error needs fix** | Auto-Fix Principles, Fix Loop Principles | [08-auto-fix-principles.md](08-auto-fix-principles.md) |
+| **Coding Implementation** | Coding Style, SOLID Principles, CEP Series | [04-operational-principles.md](04-operational-principles.md#ii-o1-coding-style) |
+| **Code Review** | Review Standards, Checklists, CRP Series | [05-domain-principles.md](05-domain-principles.md#v-code-review-principles) |
+| **Task Breakdown** | ATOMIC Principles, Task Breakdown Quality | [03-tactical-principles.md](03-tactical-principles.md#ii-t1-task-breakdown-quality-atomic-principles) |
+| **Requirement Analysis** | Structured Clarification Protocol, Requirement Layering, REP Series | [05-domain-principles.md](05-domain-principles.md#iii-requirement-engineering-principles) |
+| **Solution Design** | DDD, System Design, DDP/SDP Series | [05-domain-principles.md](05-domain-principles.md#i-ddd-tactical-design-principles) |
+| **Safety Red Lines** | Constitution Gate, Stop Conditions | [09-constitution.md](09-constitution.md) |
+| **Framework Evolution** | Three-Layer Evolution System | [10-evolution-mechanism.md](10-evolution-mechanism.md) |
+
+### Principle Quick Reference
+
+#### L1: First Principles (3 items)
+
+| Principle | One-Liner | Application Scenario |
+|-----------|-----------|---------------------|
+| P1 Verification-First | Deterministic Verification > AI Review > AI Self-Review | All output validation |
+| P2 Minimal-Viable Process | If one file suffices, don't split into two | Knowledge management, task breakdown |
+| P3 Progressive Autonomy | From "human reviews every step" to "human only manages harness" | Autonomous decision boundaries |
+
+#### L2: Strategic Principles (4 items)
+
+| Principle | One-Liner | Application Scenario |
+|-----------|-----------|---------------------|
+| S1 Two-Layer Validation | Layer 1 must pass before Layer 2 | Before completing any stage |
+| S2 Do-Review Separation | Producer and reviewer are different AI instances | When starting review tasks |
+| S3 Harness Engineering | When unsatisfied with output, change the harness | When output quality is poor |
+| S4 Informed Guessing | Maximum 3 NEEDS CLARIFICATION items | Requirement ambiguity handling |
+
+#### L3: Tactical Principles (6 items)
+
+| Principle | One-Liner | Application Scenario |
+|-----------|-----------|---------------------|
+| T1 Task Breakdown | ATOMIC principles, 2-4 hour granularity | Task decomposition |
+| T2 Context Management | Hot/Warm/Cold layering, load on demand | Knowledge loading |
+| T3 Precision Rollback | Retry only failed items, not full redo | Fix loop |
+| T4 Anti-Pattern Inoculation | Block 6 shortcut paths | Skill design |
+| T5 Taming Non-Determinism | Four defense lines, hypothesis management | Critical decisions |
+| T6 Prioritize Reuse | Search before implement, no hardcoding | Coding / fixing |
+
+#### L4: Operational Principles (4 items)
+
+| Principle | One-Liner | Application Scenario |
+|-----------|-----------|---------------------|
+| O1 Coding Style | SOLID + DDD, Lock-Check-Update | Code writing |
+| O2 Auto-Fix | Analyze→Decide→Execute→Verify, max 3 times | Error handling |
+| O3 Constitution Gate | 15 red lines, stop immediately on violation | Full process |
+| O4 Checklists | 5-stage checklists, embedded self-check | Stage completion |
+
+---
+
+## Principle Traceability Chain Examples
+
+**Scenario**: Maven dependency missing error
+
+```
+Error Occurs
+    ↓
+L4-O2 Auto-Fix Mechanism → Error Classification (Dependency Type)
+    ↓
+L3-T6 Prioritize Reusing Project Capabilities → Scan existing pom.xml
+    ↓
+L3-T3 Precision Rollback → Level 2 Fix Loop
+    ↓
+L2-S1 Two-Layer Validation → Re-execute original command for verification
+    ↓
+L1-P1 Verification-First → Verification result passed
+    ↓
+Record practice-log → L3-Evolution Mechanism
+```
+
+**Scenario**: Requirement ambiguity needs clarification
+
+```
+Requirement Ambiguity Identified
+    ↓
+L2-S4 Informed Guessing > Infinite Questions → Make reasonable default + mark
+    ↓
+L3-T5 Taming Non-Determinism → Defense Line 2: Anchor Consistency
+    ↓
+L2-Constitution #11 → Maximum 3 NEEDS CLARIFICATION items
+    ↓
+L3-T4 Anti-Pattern Inoculation → Prevent assumption drift
+    ↓
+L4-templates → Write to requirement document
+```
+
+Complete traceability chain: [00-principles-hierarchy.md](00-principles-hierarchy.md)
+
+---
+
+## Evolution Zone Description
+
+| Zone | Description | Change Method |
+|------|-------------|---------------|
+| 🔴 **Frozen** | Constitution, Evolution Rules, Core Principles, Design Philosophy | Human-only changes, requires complete review process |
+| 🟡 **Controlled** | Process Definitions, Universal Skills, Checklists, Coding Standards | Auto-proposed + Human-approved |
+| 🟢 **Free** | Domain Skills, Benchmarks, Support Skills | Auto-evolved + Validator-verified |
+| 🔵 **Environment Knowledge** | Middleware, Components, Integration Knowledge | Project-level auto-optimization |
+
+---
+
+## Mapping to Related Documents
+
+| Document Section | Corresponding Document | Description |
+|-----------------|----------------------|-------------|
+| L1 First Principles | [01-first-principles.md](01-first-principles.md) | Triangular Prism P1/P2/P3 |
+| L2 Strategic Principles | [02-strategic-principles.md](02-strategic-principles.md) | S1-S4 |
+| L3 Tactical Principles | [03-tactical-principles.md](03-tactical-principles.md) | T1-T6 |
+| L4 Operational Principles | [04-operational-principles.md](04-operational-principles.md) | O1-O4 |
+| Domain Principles | [05-domain-principles.md](05-domain-principles.md) | DDP/SDP/REP/CEP/CRP/KMP |
+| Methodology Principles | [06-methodology-principles.md](06-methodology-principles.md) | P1-P44 |
+| Cognitive Enhancement Principles | [07-cognitive-principles.md](07-cognitive-principles.md) | CP1-CP10 |
+| Auto-Fix Principles | [08-auto-fix-principles.md](08-auto-fix-principles.md) | AFP1-AFP6 |
+| Project Constitution | [09-constitution.md](09-constitution.md) | 15 Red Lines |
+| Evolution Mechanism | [10-evolution-mechanism.md](10-evolution-mechanism.md) | Three-Layer Self-Evolution System |
+
+---
+
+## Usage Guide
+
+### Quick Principle Location
+
+**Scenario**: What principle should I follow for problem X?
+
+1. Identify problem type (Verification / Autonomy / Knowledge / Fix / ...)
+2. Check the Principle Quick Reference Table
+3. Trace upstream principles (understand why)
+4. Check downstream operational standards (know how)
+
+**Example**:
+```
+Problem: "AI always reinvents the wheel, doesn't scan existing code"
+→ Problem Type: Knowledge Management / Reuse
+→ Quick Reference: T6 Prioritize Reusing Project Capabilities
+→ Upstream Principles: P2 Minimal-Viable, S3 Harness Engineering
+→ Downstream Operations: 04-operational-principles.md Coding Standards, 08-auto-fix-principles.md AFP3
+```
+
+### Principle Application Checkpoints
+
+**Before Executing Task**:
+- [ ] Read this document, understand relevant principles
+- [ ] Identified task type and applicable principles
+- [ ] Prepared necessary knowledge (Hot/Warm layers)
+
+**During Task Execution**:
+- [ ] Decompose tasks following ATOMIC principles
+- [ ] Follow Informed Guessing principle (max 3 uncertain items)
+- [ ] Follow Do-Review Separation principle (start independent Agent for review)
+
+**After Task Execution**:
+- [ ] Follow Two-Layer Validation principle (Layer 1 + Layer 2)
+- [ ] Follow Checklist principle (stage self-check)
+- [ ] Follow Practice Log principle (record experience)
+
+---
+
+**Maintainer**: AutoSpec Team
+**Feedback Channel**: Knowledge Base Git Repository Issue
+**Related Documents**:
+- [Principles Hierarchy](00-principles-hierarchy.md)
+- [L1 First Principles](01-first-principles.md)
+- [L2 Strategic Principles](02-strategic-principles.md)
+- [L3 Tactical Principles](03-tactical-principles.md)
+- [L4 Operational Principles](04-operational-principles.md)
+- [Domain Practice Principles](05-domain-principles.md)
+- [Methodology Principles](06-methodology-principles.md)
+- [Cognitive Enhancement Principles](07-cognitive-principles.md)
+- [Auto-Fix Principles](08-auto-fix-principles.md)
+- [Project Constitution](09-constitution.md)
+- [Self-Evolution Mechanism](10-evolution-mechanism.md)