| name | pdd-implement-feature |
| description | 根据开发规格实现功能点代码的核心Skill。当用户想要开始编码实现、根据规格文档生成代码、实现功能点时调用此Skill。此Skill会自动调用pdd-template-engine生成基础代码框架,然后由software-engineer补充业务逻辑。即使只有规格文档没有明确说'实现',只要涉及代码生成、功能开发,都应触发此Skill。支持中文触发:实现功能点、编码实现、开始编码、功能开发、代码实现、PDD实现。 |
| license | MIT |
| compatibility | Requires specification generation to be completed first |
| metadata | {"author":"neuqik@hotmail.com","version":"2.0","parent":"pdd-main"} |
Feature Point Implementation - 基于开发规格的功能点实现 / Feature Point Code Implementation Based on Development Specifications
输入: spec.md(开发规格) | checklist.md(验收标准) | test-cases.md(测试用例,可选)
Input: spec.md (Development Spec) | checklist.md (Acceptance Criteria) | test-cases.md (Test Cases, Optional)
输出: 代码文件 | 验收报告
Output: Code Files | Acceptance Report
技能集成 / Skill Integration
🇨🇳 Software Engineer 调用
| 调用时机 | 服务内容 |
|---|
| 代码实现 | 基于规格执行代码实现 |
| 单元测试 | 编写单元测试和集成测试 |
| 代码重构 | 按编码规范优化代码 |
| 缺陷修复 | 错误处理和问题解决 |
🇺🇸 Software Engineer Invocation
| Trigger | Service |
|---|
| Code Implementation | Execute code implementation based on specifications |
| Unit Testing | Write unit tests and integration tests |
| Code Refactoring | Optimize code per coding standards |
| Defect Fixing | Error handling and issue resolution |
🇨🇳 Expert Skill 调用(按需)
| 专家技能 | 触发条件 | 期望输出 |
|---|
| expert-ruoyi | 若依框架问题 | 解决方案+最佳实践 |
| expert-activiti | 工作流问题 | BPMN设计建议 |
| expert-mysql | 数据库问题 | SQL优化方案 |
| expert-code-quality | 代码质量问题 | 重构方案 |
🇺🇸 Expert Skill Invocation (On-Demand)
| Expert Skill | Trigger Condition | Expected Output |
|---|
| expert-ruoyi | RuoYi framework issues | Solution + Best Practices |
| expert-activiti | Workflow issues | BPMN design recommendations |
| expert-mysql | Database issues | SQL optimization plan |
| expert-code-quality | Code quality issues | Refactoring plan |
流程步骤 / Process Steps
🇨🇳 Step 1: 读取开发规格
从 dev-specs/FP-{序号}/spec.md 读取: 接口定义 | 数据模型 | 业务逻辑 | 测试用例
🇺🇸 Step 1: Read Development Specification
Read from dev-specs/FP-{sequence}/spec.md: API Definitions | Data Models | Business Logic | Test Cases
🇨🇳 Step 1.5: 上下文注入(关键改进)
在生成代码前,扫描项目已有代码,构建项目上下文,避免AI孤岛效应:
- 扫描已有Model: 从
models/ 目录提取所有模型类名和字段列表
- 扫描已有路由: 从
api/ 目录提取所有已注册路由
- 扫描已有Schema: 从
schemas/ 目录提取所有Schema定义
- 读取Bug模式库: 从
config/bug-patterns.yaml 提取Bug模式
- 读取命名约定: 从PRD/Spec提取枚举编码、类型映射等约定
- 扫描前后端契约关系: 从后端 Controller 提取所有 @RequestMapping 路径和方法签名,从前端 api/ 目录提取所有 request() 调用的 URL,建立 {后端路径 → 前端调用文件} 的映射表
- 扫描状态映射关系: 全局搜索所有包含 getStatusLabel / statusMap / getNodeLabel / getStatusTagType 等方法的文件,建立 {状态值 → 使用文件列表} 的映射表
将以上信息作为额外上下文注入到代码生成过程中,确保:
- 新代码正确import已有Model(防止PATTERN-C4: import缺失)
- 新路由不与已有路由冲突(防止PATTERN-002: 路由顺序错误)
- 新Schema与已有Schema字段一致(防止PATTERN-C3: 字段不同步)
- 遵守Bug模式库中的约束(防止重复犯错)
- 新代码的前端API路径与后端Controller完整路径一致(防止PATTERN-R008: API路径断层)
- 新增状态值时所有映射文件同步更新(防止PATTERN-R011: 状态字典不完整)
🇺🇸 Step 1.5: Context Injection (Critical Improvement)
Before generating code, scan existing project code to build project context, avoiding AI island effect:
- Scan existing Models: Extract all model class names and field lists from
models/ directory
- Scan existing Routes: Extract all registered routes from
api/ directory
- Scan existing Schemas: Extract all Schema definitions from
schemas/ directory
- Read Bug Pattern Library: Extract bug patterns from
config/bug-patterns.yaml
- Read Naming Conventions: Extract enum coding, type mapping conventions from PRD/Spec
- Scan Frontend-Backend Contract Relations: Extract all @RequestMapping paths and method signatures from backend Controllers, extract all request() call URLs from frontend api/ directory, build {backend path → frontend call file} mapping table
- Scan Status Mapping Relations: Global search all files containing getStatusLabel / statusMap / getNodeLabel / getStatusTagType methods, build {status value → usage file list} mapping table
Inject the above information as additional context into the code generation process, ensuring:
- New code correctly imports existing Models (prevent PATTERN-C4: missing import)
- New routes don't conflict with existing routes (prevent PATTERN-002: route order errors)
- New Schemas are consistent with existing Schemas (prevent PATTERN-C3: field mismatch)
- Bug pattern library constraints are followed (prevent recurring bugs)
- New frontend API paths match backend Controller complete paths (prevent PATTERN-R008: API path mismatch)
- When adding new status values, all mapping files are synchronized (prevent PATTERN-R011: incomplete status dictionary)
🇨🇳 Step 2: 读取验收标准
从 dev-specs/FP-{序号}/checklist.md 读取验收项
🇺🇸 Step 2: Read Acceptance Criteria
Read acceptance items from dev-specs/FP-{sequence}/checklist.md
🇨🇳 Step 3: 确定实现顺序
依赖顺序: 数据模型→数据库脚本 → 后端接口(Controller/Service/Mapper) → 前端页面(Vue组件)
🇺🇸 Step 3: Determine Implementation Order
Dependency order: Data Model → Database Scripts → Backend APIs (Controller/Service/Mapper) → Frontend Pages (Vue Components)
🇨🇳 Step 4: 生成数据库脚本
基于数据模型生成SQL(含审计字段: create_time/update_time/create_by/update_by/del_flag/status)
🇺🇸 Step 4: Generate Database Scripts
Generate SQL based on data model (including audit fields: create_time/update_time/create_by/update_by/del_flag/status)
🇨🇳 Step 5: 生成后端代码(委托 software-engineer)
Domain实体类(@Data/@TableName) | Mapper接口(@Mapper/BaseMapper) | Service接口和实现(@Service/ServiceImpl) | Controller(@RestController/@RequestMapping)
🇺🇸 Step 5: Generate Backend Code (Delegate to software-engineer)
Domain Entity Classes (@Data/@TableName) | Mapper Interfaces (@Mapper/BaseMapper) | Service Interface & Implementation (@Service/ServiceImpl) | Controller (@RestController/@RequestMapping)
🇨🇳 Step 6: 生成前端代码(委托 software-engineer)
API接口(request封装) | Vue组件(template/script/data/methods)
🇺🇸 Step 6: Generate Frontend Code (Delegate to software-engineer)
API Interfaces (request wrapper) | Vue Components (template/script/data/methods)
🇨🇳 Step 7: 实现业务逻辑
处理流程 | 校验规则 | 状态转换 | 异常处理
专家咨询(按需): RuoYi问题→expert-ruoyi | 数据库问题→expert-mysql
🇺🇸 Step 7: Implement Business Logic
Process Flow | Validation Rules | State Transitions | Exception Handling
Expert Consultation (On-Demand): RuoYi Issues → expert-ruoyi | Database Issues → expert-mysql
🇨🇳 Step 8: 微验证(关键改进)
每完成一个功能点的代码实现后,必须立即执行微验证(约30秒),不要等所有功能点写完再测试:
- 后端启动检查:
python -m uvicorn app.main:app --reload 无报错
- API端点可达: 访问
GET /docs 能看到Swagger文档
- Schema序列化: 创建+查询各1次,确认datetime/Enum序列化正确
- 前端编译检查:
npm run build 无报错
- Options接口验证:
curl GET /{module}/options 返回非空数据
微验证失败时,必须立即修复,不得继续下一个功能点。
🇺🇸 Step 8: Micro-Verification (Critical Improvement)
After completing each feature point's code implementation, immediately execute micro-verification (~30 seconds). Do not wait until all feature points are written:
- Backend startup check:
python -m uvicorn app.main:app --reload no errors
- API endpoint reachable: Access
GET /docs to see Swagger documentation
- Schema serialization: Create + query once each, confirm datetime/Enum serialization is correct
- Frontend compilation check:
npm run build no errors
- Options API verification:
curl GET /{module}/options returns non-empty data
When micro-verification fails, fix immediately. Do not proceed to the next feature point.
🇨🇳 Step 9: 运行完整测试
单元测试 | 接口测试 | 集成测试
🇺🇸 Step 9: Run Full Tests
Unit Tests | API Tests | Integration Tests
🇨🇳 Step 10: 更新验收状态
更新checklist.md中的验收状态
🇺🇸 Step 10: Update Acceptance Status
Update acceptance status in checklist.md
🇨🇳 Step 11: 生成验收报告
输出: 业务验收 | 技术验收 | 问题日志 | 结论(通过/不通过)
🇺🇸 Step 11: Generate Acceptance Report
Output: Business Acceptance | Technical Acceptance | Issue Log | Conclusion (Pass/Fail)
代码标准 / Code Standards
🇨🇳 后端标准
类名PascalCase | 方法名camelCase | 常量UPPER_SNAKE_CASE | 注释Javadoc格式
🇺🇸 Backend Standards
Class names PascalCase | Method names camelCase | Constants UPPER_SNAKE_CASE | Comments in Javadoc format
🇨🇳 前端标准
组件名PascalCase | 方法名camelCase | CSS类名kebab-case | ES6+语法
🇺🇸 Frontend Standards
Component names PascalCase | Method names camelCase | CSS class names kebab-case | ES6+ syntax
🇨🇳 software-engineer 标准
先读现有代码风格再写新代码 | 错误处理优先 | 保持最小化 | PR-ready代码
🇺🇸 software-engineer Standards
Read existing code style before writing new code | Error handling first | Keep it minimal | PR-ready code
PR管理提示(功能点完成后)/ PR Management Tips (After Feature Point Completion)
🇨🇳
完成并验证通过后,提示用户可手动调用PR管理技能:
/pdd-pr-create {change-id} - 创建PR并执行自动审查
/pdd-pr-review {change-id} - 查看PR审查结果
/pdd-pr-merge {change-id} - 合并PR并归档
重要: PDD框架不会自动调用 pdd-pr-* 技能,需用户手动决定是否使用PR管理功能。
🇺🇸
After completion and verification, prompt the user to manually invoke PR management skills:
/pdd-pr-create {change-id} - Create PR and execute auto-review
/pdd-pr-review {change-id} - View PR review results
/pdd-pr-merge {change-id} - Merge PR and archive
Important: The PDD framework will NOT automatically call pdd-pr-* skills. Users must manually decide whether to use PR management features.
错误处理与回退规范 / Error Handling & Rollback Rules
🇨🇳 错误分级
Critical(必须修复,阻塞流程) | Warning(建议修复,非阻塞) | Suggestion(可选优化)
🇺🇸 Error Classification
Critical (Must fix, blocks process) | Warning (Recommended fix, non-blocking) | Suggestion (Optional optimization)
🇨🇳 重试策略
限制: 每个功能点最多3次 | 超限: 暂停流程,等待人工决策
🇺🇸 Retry Strategy
Limit: Max 3 retries per feature point | Exceeded: Pause process, await human decision
🇨🇳 回退规则
pdd-code-reviewer审查失败 → 返回pdd-implement-feature重新实现 | pdd-verify-feature验证失败 → 返回pdd-implement-feature重新验证
🇺🇸 Rollback Rules
pdd-code-reviewer review failed → Return to pdd-implement-feature for re-implementation | pdd-verify-feature verification failed → Return to pdd-implement-feature for re-verification
🇨🇳 失败记录
位置: dev-specs/FP-{模块}-{序号}/review-report.md
内容: 失败时间/阶段/原因/尝试次数/相关错误日志
🇺🇸 Failure Record
Location: dev-specs/FP-{module}-{sequence}/review-report.md
Content: Failure time / Stage / Reason / Attempt count / Related error logs
Guardrails
🇨🇳
- 代码必须符合项目标准
- 必须实现规格中定义的所有接口
- 必须处理规格中定义的所有异常
- 必须通过所有验收项才能标记完成
- 代码变更后必须同步更新规格文档
- 遇到技术问题时必须咨询专家技能
- 代码实现必须遵循software-engineer标准
- 功能点完成后提示用户可用PR管理技能,但不自动调用
🇺🇸
- Code must comply with project standards
- All interfaces defined in the specification must be implemented
- All exceptions defined in the specification must be handled
- Must pass all acceptance items before marking as complete
- Specification documents must be updated synchronously after code changes
- Must consult expert skills when encountering technical issues
- Code implementation must follow software-engineer standards
- After feature point completion, prompt user about PR management skills but do not auto-invoke
Iron Law / 铁律
🇨🇳
-
规格即法律: 代码实现必须严格遵循spec.md的定义,不得擅自添加规格中未定义的功能,也不得遗漏规格要求的任何接口或字段。
-
委托不替代: 遇到框架特定问题时,必须调用对应的expert技能获取方案,不得凭猜测或通用知识硬编码。
-
错误处理优先: 业务逻辑实现前,必须先完成参数校验、异常处理、边界条件等防御性代码,不得只写"快乐路径"。
-
代码标准对齐: 生成的代码风格必须与项目现有代码保持一致,编写前先阅读同类模块的现有代码。
-
验收驱动开发: 每完成一个接口实现,立即对照checklist.md进行自检,不要等到全部写完再统一检查。
-
上下文感知生成: 生成代码前必须扫描项目已有代码,了解已有Model/Schema/路由,避免孤岛效应导致的import缺失、字段不同步、路由冲突。
-
微验证即时反馈: 每完成一个功能点必须立即执行微验证(后端启动+API可达+Schema序列化+前端编译),不得跳过微验证继续下一个功能点。
-
Bug模式库约束: 代码生成必须遵守 config/bug-patterns.yaml 中的Bug模式库,已知陷阱不得重复触犯。
违规示例: ❌ 实现了规格中未定义的功能(功能蔓延) | ❌ 遇到若依权限注解问题时不调用expert-ruoyi | ❌ 只实现正常流程未处理异常情况 | ❌ 使用与项目不同的命名风格 | ❌ 写完所有代码后才运行发现编译错误 | ❌ 不扫描已有代码直接生成新代码(孤岛效应) | ❌ 跳过微验证直接进入下一个功能点 | ❌ datetime字段声明为str而非datetime(PATTERN-001) | ❌ /options路由放在/{id}之后(PATTERN-002)
合规示例: ✅ 严格按照spec.md的接口列表逐个实现 | ✅ 遇到@PreAuthorize问题时调用expert-ruoyi | ✅ 每个Service方法开头进行参数校验 | ✅ 先阅读现有TransferApplyController保持相同风格 | ✅ 每完成一个Controller方法立即测试 | ✅ 生成代码前扫描models/和schemas/目录 | ✅ 每个功能点完成后执行微验证 | ✅ datetime字段使用field_serializer(PATTERN-001) | ✅ /options路由在/{id}之前注册(PATTERN-002)
🇺🇸
-
Specification is Law: Code implementation must strictly follow the definitions in spec.md. Do not arbitrarily add features not defined in the specification, nor omit any interfaces or fields required by the spec.
-
Delegate, Don't Substitute: When encountering framework-specific issues, you must invoke the corresponding expert skill to obtain a solution. Do not hard-code based on guesses or generic knowledge.
-
Error Handling First: Before implementing business logic, you must complete defensive code including parameter validation, exception handling, and boundary conditions. Do not only write the "happy path".
-
Code Standard Alignment: The generated code style must be consistent with the existing project code. Read the existing code of similar modules before writing new code.
-
Acceptance-Driven Development: After completing each interface implementation, immediately self-check against checklist.md. Do not wait until all code is written to perform a unified check.
-
Context-Aware Generation: Before generating code, you must scan existing project code to understand existing Models/Schemas/Routes, avoiding the island effect that causes missing imports, field mismatches, and route conflicts.
-
Micro-Verification Immediate Feedback: After completing each feature point, immediately execute micro-verification (backend startup + API reachable + Schema serialization + frontend compilation). Do not skip micro-verification to proceed to the next feature point.
-
Bug Pattern Library Constraints: Code generation must comply with the Bug Pattern Library in config/bug-patterns.yaml. Known pitfalls must not be repeatedly violated.
Violation Examples: ❌ Implemented features not defined in the spec (feature creep) | ❌ Did not invoke expert-ruoyi when encountering RuoYi permission annotation issues | ❌ Only implemented normal flow without handling exceptions | ❌ Used naming conventions different from the project | ❌ Ran tests only after writing all code and discovered compilation errors | ❌ Generating code without scanning existing code (island effect) | ❌ Skipping micro-verification and proceeding to next feature point | ❌ Declaring datetime field as str instead of datetime (PATTERN-001) | ❌ Registering /options route after /{id} (PATTERN-002)
Compliance Examples: ✅ Implemented each interface one by one strictly following the spec.md interface list | ✅ Invoked expert-ruoyi when encountering @PreAuthorize issues | ✅ Performed parameter validation at the beginning of each Service method | ✅ Read existing TransferApplyController first to maintain consistent style | ✅ Tested immediately after completing each Controller method | ✅ Scanning models/ and schemas/ directories before generating code | ✅ Executing micro-verification after each feature point completion | ✅ Using field_serializer for datetime fields (PATTERN-001) | ✅ Registering /options route before /{id} (PATTERN-002)
Rationalization Table / 理性化对照表
🇨🇳
| # | 你可能的想法 / Trap | 请问自己 / Question | 应该怎么做 / Action |
|---|
| 1 | "这个功能用户可能需要,顺便加上吧" | 未在规格中的功能是范围蔓延,会导致验收失败 | 严格按规格实现,额外需求走正式变更流程 |
| 2 | "这个框架问题我以前遇到过" | 框架版本差异可能导致旧方案失效 | 直接调用对应expert技能获取经过验证的解决方案 |
| 3 | "异常处理后面统一加" | 异常处理后置往往意味着永远不会加 | 在写业务逻辑前先搭建好异常处理骨架 |
| 4 | "项目代码风格太乱了,我按我的习惯写吧" | 不一致的代码风格会增加维护负担 | 先阅读现有代码,适配项目既有风格 |
| 5 | "先全部写完再一起测试" | 大批量代码调试难度远高于增量验证 | 采用TDD思维:实现一个就验证一个 |
常见陷阱:
- "规格偏离"陷阱: 实现过程中逐渐偏离规格要求 → 建立实现清单,每完成一项打勾并与规格对照
- "框架硬扛"陷阱: 遇到框架问题反复尝试而不求助专家 → 设定"3次尝试规则",3次未解决必须调用expert
- "快乐路径"陷阱: 只实现正常流程忽略异常处理 → 强制异常覆盖率:每个public方法必须有try-catch
- "风格割裂"陷阱: 新代码与项目现有代码风格不一致 → 实现"先读后写"规则
🇺🇸
| # | You Might Think / Trap | Ask Yourself / Question | What To Do / Action |
|---|
| 1 | "Users might need this feature, let me add it too" | Features not in the spec are scope creep and will cause acceptance failure | Implement strictly per spec; extra requirements go through formal change process |
| 2 | "I've encountered this framework issue before" | Framework version differences may render old solutions invalid | Invoke the corresponding expert skill to get a verified solution |
| 3 | "I'll add exception handling later" | Postponed exception handling often means it never gets added | Build the exception handling skeleton before writing business logic |
| 4 | "The project code style is messy, I'll write in my own style" | Inconsistent code style increases maintenance burden | Read existing code first and adapt to project conventions |
| 5 | "I'll test everything after writing all the code" | Debugging a large batch of code is far harder than incremental verification | Adopt TDD mindset: implement one, verify one |
Common Traps:
- "Spec Drift" Trap: Gradually deviating from spec requirements during implementation → Create an implementation checklist; check off each item against the spec
- "Framework Stubbornness" Trap: Repeatedly trying to solve framework issues without consulting experts → Enforce "3-attempt rule": must invoke expert after 3 failed attempts
- "Happy Path" Trap: Only implementing normal flow while ignoring exception handling → Enforce exception coverage: every public method must have try-catch
- "Style Fragmentation" Trap: New code inconsistent with existing project code style → Implement "read-first, write-later" rule
Red Flags / 红旗警告
Layer 1: 输入检查 / Input Validation Guards
🇨🇳
- INPUT-IMPL-001: spec.md或checklist.md不存在 → 🔴 CRITICAL → 终止并提示先生成开发规格
- INPUT-IMPL-002: spec.md缺少接口定义或数据模型章节 → 🔴 CRITICAL → 提示规格文档不完整
- INPUT-IMPL-003: 项目路径不存在或无法访问 → 🔴 CRITICAL → 终止并检查项目路径配置
🇺🇸
- INPUT-IMPL-001: spec.md or checklist.md does not exist → 🔴 CRITICAL → Terminate and prompt to generate development specification first
- INPUT-IMPL-002: spec.md is missing API definitions or data model sections → 🔴 CRITICAL → Prompt that specification document is incomplete
- INPUT-IMPL-003: Project path does not exist or is inaccessible → 🔴 CRITICAL → Terminate and check project path configuration
Layer 2: 执行检查 / Execution Guards
🇨🇳
- EXEC-IMPL-001: 实现了规格中未定义的接口或字段 → 🟡 WARN → 标记为"超出规格",请用户确认
- EXEC-IMPL-002: 遇到框架相关问题但未调用expert技能而自行硬编码 → 🔴 CRITICAL → 回退修改,调用expert获取正确方案
- EXEC-IMPL-003: 代码存在空指针风险(未做null检查) → 🟡 WARN → 补充null检查或Optional
- EXEC-IMPL-004: SQL语句存在注入风险(字符串拼接) → 🔴 CRITICAL → 必须使用参数化查询
- EXEC-IMPL-005: 生成代码前未扫描项目已有代码(未执行Step 1.5上下文注入) → 🔴 CRITICAL → 回退执行上下文注入步骤
- EXEC-IMPL-006: 功能点完成后跳过微验证(未执行Step 8) → 🔴 CRITICAL → 立即执行微验证,修复问题后才能继续
🇺🇸
- EXEC-IMPL-001: Implemented interfaces or fields not defined in the specification → 🟡 WARN → Mark as "out of scope" and request user confirmation
- EXEC-IMPL-002: Encountered framework-related issues but hard-coded without invoking expert skills → 🔴 CRITICAL → Rollback changes, invoke expert for correct solution
- EXEC-IMPL-003: Code has null pointer risk (no null checks performed) → 🟡 WARN → Add null checks or use Optional
- EXEC-IMPL-004: SQL statements have injection risk (string concatenation) → 🔴 CRITICAL → Must use parameterized queries
- EXEC-IMPL-005: Did not scan existing project code before generating code (Step 1.5 context injection not executed) → 🔴 CRITICAL → Rollback and execute context injection step
- EXEC-IMPL-006: Skipped micro-verification after feature point completion (Step 8 not executed) → 🔴 CRITICAL → Execute micro-verification immediately, fix issues before continuing
Layer 3: 输出检查 / Output Validation Guards
🇨🇳
- OUTPUT-IMPL-001: 代码文件有语法错误(无法编译) → 🔴 CRITICAL → 修复后再标记完成
- OUTPUT-IMPL-002: 规格要求的接口未全部实现 → 🔴 CRITICAL → 补充遗漏的实现
- OUTPUT-IMPL-003: 代码文件路径不符合规范(如放在asset-system下) → 🔴 CRITICAL → 移动到正确路径
- OUTPUT-IMPL-004: datetime字段声明为str而非datetime类型(违反PATTERN-001) → 🔴 CRITICAL → 修改为datetime类型并添加field_serializer
- OUTPUT-IMPL-005: /options路由注册在/{id}路由之后(违反PATTERN-002) → 🔴 CRITICAL → 调整路由注册顺序,/options必须在/{id}之前
🇺🇸
- OUTPUT-IMPL-001: Code files have syntax errors (cannot compile) → 🔴 CRITICAL → Fix before marking as complete
- OUTPUT-IMPL-002: Not all interfaces required by the specification are implemented → 🔴 CRITICAL → Supplement missing implementations
- OUTPUT-IMPL-003: Code file path does not conform to standards (e.g., placed under asset-system) → 🔴 CRITICAL → Move to correct path
- OUTPUT-IMPL-004: datetime field declared as str instead of datetime type (violates PATTERN-001) → 🔴 CRITICAL → Change to datetime type and add field_serializer
- OUTPUT-IMPL-005: /options route registered after /{id} route (violates PATTERN-002) → 🔴 CRITICAL → Adjust route registration order, /options must be before /{id}
触发Red Flag时的处理流程 / Red Flag Trigger Handling
🇨🇳
🔴 CRITICAL → 立即停止,报告问题详情,等待指示 | 🟡 WARN → 记录警告,尝试自动修复,在报告中标注 | 🔵 INFO → 记录信息,正常继续
🇺🇸
🔴 CRITICAL → Stop immediately, report issue details, await instructions | 🟡 WARN → Record warning, attempt auto-fix, annotate in report | 🔵 INFO → Log information, continue normally