builder-role-skill
Compare original and translation side by side
🇺🇸
Original
English🇨🇳
Translation
ChineseBuilder Role Skill
Builder角色技能
Description
描述
Implement production-quality features, components, and systems using test-driven development, incremental delivery, and professional software engineering practices.
采用测试驱动开发(TDD)、增量交付以及专业软件工程实践,实现生产级别的功能、组件与系统。
When to Use This Skill
何时使用该技能
- Implementing features from architectural specifications
- Building new components or systems
- Converting designs into production code
- Feature development requiring multi-phase implementation
- TDD-based development workflows
- 根据架构规范实现功能
- 构建新组件或系统
- 将设计转化为生产代码
- 需要多阶段实现的功能开发
- 基于TDD的开发工作流
When NOT to Use This Skill
何时不使用该技能
- For system design or architecture (use architect-role-skill)
- For testing and code review (use validator-role-skill)
- For documentation generation (use scribe-role-skill)
- For infrastructure work (use devops-role-skill)
- 系统设计或架构工作(使用architect-role-skill)
- 测试与代码评审工作(使用validator-role-skill)
- 文档生成工作(使用scribe-role-skill)
- 基础设施工作(使用devops-role-skill)
Prerequisites
前置条件
- Architecture or design specification available
- Development environment configured
- Test framework in place
- Git repository initialized
- 具备架构或设计规范文档
- 已配置开发环境
- 已搭建测试框架
- 已初始化Git仓库
Workflow
工作流
Phase 1: Task Decomposition
阶段1:任务拆分
Break architectural specifications into granular, implementable tasks.
Step 1.1: Load Planning Documents
Read context files:
- DEVELOPMENT_PLAN.md (or equivalent)
- ARCHITECTURE.md (design specifications)
- TODO.md (task backlog)
- MULTI_AGENT_PLAN.md (if multi-agent workflow)Step 1.2: Create Implementation Plan
Generate with structured task breakdown:
IMPLEMENTATION_PLAN.mdmarkdown
undefined将架构规范拆解为粒度精细、可落地的任务。
步骤1.1:加载规划文档
Read context files:
- DEVELOPMENT_PLAN.md (or equivalent)
- ARCHITECTURE.md (design specifications)
- TODO.md (task backlog)
- MULTI_AGENT_PLAN.md (if multi-agent workflow)步骤1.2:创建实现计划
生成结构化任务拆解的:
IMPLEMENTATION_PLAN.mdmarkdown
undefinedImplementation Plan: [Feature Name]
Implementation Plan: [Feature Name]
Architecture Reference
Architecture Reference
- Source: [Document] Section [X]
- Components: [List components to build]
- Dependencies: [External/internal dependencies]
- Source: [Document] Section [X]
- Components: [List components to build]
- Dependencies: [External/internal dependencies]
Phase Breakdown
Phase Breakdown
Phase 1: Foundation
Phase 1: Foundation
- Task 1.1: [Specific task]
- Task 1.2: [Specific task]
- Task 1.3: Write unit tests for foundation
- Task 1.1: [Specific task]
- Task 1.2: [Specific task]
- Task 1.3: Write unit tests for foundation
Phase 2: Business Logic
Phase 2: Business Logic
- Task 2.1: [Specific task]
- Task 2.2: [Specific task]
- Task 2.3: Write service tests
- Task 2.1: [Specific task]
- Task 2.2: [Specific task]
- Task 2.3: Write service tests
Phase 3: Integration
Phase 3: Integration
- Task 3.1: [Specific task]
- Task 3.2: Write integration tests
- Task 3.3: Full test suite validation
- Task 3.1: [Specific task]
- Task 3.2: Write integration tests
- Task 3.3: Full test suite validation
Completion Criteria
Completion Criteria
- All tests passing (coverage >= 80%)
- Code linted with zero errors
- Documentation updated
- Peer review approved
**Step 1.3: Validate Plan**
- Ensure tasks are specific and measurable
- Verify dependencies are identified
- Confirm test coverage is planned
- Check phase estimates are realistic
---- All tests passing (coverage >= 80%)
- Code linted with zero errors
- Documentation updated
- Peer review approved
**步骤1.3:验证计划**
- 确保任务具体且可衡量
- 确认已识别所有依赖
- 确认已规划测试覆盖范围
- 检查阶段预估是否合理
---Phase 2: TDD Implementation Loop
阶段2:TDD实现循环
For each task in the implementation plan, follow the RED-GREEN-REFACTOR cycle:
Step 2.1: RED - Write Failing Test
javascript
// Example: test/user-service.test.js
describe('UserService', () => {
test('createUser should validate email format', async () => {
const userService = new UserService();
await expect(userService.createUser({
email: 'invalid-email',
name: 'John Doe'
})).rejects.toThrow('Invalid email format');
});
});Run test to confirm it fails:
bash
npm test -- --testPathPattern=user-service针对实现计划中的每个任务,遵循RED-GREEN-REFACTOR循环:
步骤2.1:RED - 编写失败的测试用例
javascript
// Example: test/user-service.test.js
describe('UserService', () => {
test('createUser should validate email format', async () => {
const userService = new UserService();
await expect(userService.createUser({
email: 'invalid-email',
name: 'John Doe'
})).rejects.toThrow('Invalid email format');
});
});运行测试以确认失败:
bash
npm test -- --testPathPattern=user-serviceExpected: FAIL - Test should fail initially
Expected: FAIL - Test should fail initially
**Step 2.2: GREEN - Write Minimal Implementation**
```javascript
// Example: src/services/user-service.js
class UserService {
async createUser(userData) {
// Minimal code to make test pass
const emailRegex = /^[^\s@]+@[^\s@]+\.[^\s@]+$/;
if (!emailRegex.test(userData.email)) {
throw new Error('Invalid email format');
}
// ... rest of implementation
}
}Run test to confirm it passes:
bash
npm test -- --testPathPattern=user-service
**步骤2.2:GREEN - 编写最小化实现代码**
```javascript
// Example: src/services/user-service.js
class UserService {
async createUser(userData) {
// Minimal code to make test pass
const emailRegex = /^[^\s@]+@[^\s@]+\.[^\s@]+$/;
if (!emailRegex.test(userData.email)) {
throw new Error('Invalid email format');
}
// ... rest of implementation
}
}运行测试以确认通过:
bash
npm test -- --testPathPattern=user-serviceExpected: PASS - Test should now pass
Expected: PASS - Test should now pass
**Step 2.3: REFACTOR - Improve Code Quality**
1. **Extract reusable functions**:
```javascript
// utils/validators.js
export const isValidEmail = (email) => {
const emailRegex = /^[^\s@]+@[^\s@]+\.[^\s@]+$/;
return emailRegex.test(email);
};
// Refactored service
import { isValidEmail } from '../utils/validators.js';
class UserService {
async createUser(userData) {
if (!isValidEmail(userData.email)) {
throw new Error('Invalid email format');
}
// ... rest of implementation
}
}- Run linter and formatter:
bash
npm run lint # Fix linting issues
npm run format # Auto-format code- Re-run tests:
bash
npm test
**步骤2.3:REFACTOR - 提升代码质量**
1. **提取可复用函数**:
```javascript
// utils/validators.js
export const isValidEmail = (email) => {
const emailRegex = /^[^\s@]+@[^\s@]+\.[^\s@]+$/;
return emailRegex.test(email);
};
// Refactored service
import { isValidEmail } from '../utils/validators.js';
class UserService {
async createUser(userData) {
if (!isValidEmail(userData.email)) {
throw new Error('Invalid email format');
}
// ... rest of implementation
}
}- 运行代码检查与格式化工具:
bash
npm run lint # Fix linting issues
npm run format # Auto-format code- 重新运行测试:
bash
npm testExpected: All tests still pass after refactoring
Expected: All tests still pass after refactoring
**Step 2.4: Commit Progress**
```bash
git add .
git commit -m "feat(user): add email validation to user creation
- Implement email format validation
- Add unit tests for validation
- Extract reusable validator utility
Tests: 12 passing, coverage 85%
Refs: #[issue-number]"
**步骤2.4:提交进度**
```bash
git add .
git commit -m "feat(user): add email validation to user creation
- Implement email format validation
- Add unit tests for validation
- Extract reusable validator utility
Tests: 12 passing, coverage 85%
Refs: #[issue-number]"Phase 3: Phase Integration
阶段3:阶段集成
After completing all tasks in a phase:
Step 3.1: Run Full Test Suite
bash
undefined完成某一阶段的所有任务后:
步骤3.1:运行完整测试套件
bash
undefinedRun all tests
Run all tests
npm test
npm test
Check coverage
Check coverage
npm run coverage
npm run coverage
Ensure coverage >= 80%
Ensure coverage >= 80%
Run linter
Run linter
npm run lint
npm run lint
Zero errors required
Zero errors required
**Step 3.2: Update Progress Tracking**
Mark phase complete in planning documents:
- `IMPLEMENTATION_PLAN.md` - Check off phase tasks
- `TODO.md` - Update task status
- `MULTI_AGENT_PLAN.md` - Update coordination status (if applicable)
**Step 3.3: Create Phase Commit**
```bash
git add .
git commit -m "feat([feature-name]): Complete Phase [N] - [Phase Name]
Phase [N] Summary:
- ✅ Task 1 completed
- ✅ Task 2 completed
- ✅ Task 3 completed
Tests: [X] passing, coverage [Y%]
Phase: [N] of [Total]
Next: Phase [N+1]"
**步骤3.2:更新进度跟踪**
在规划文档中标记阶段完成:
- `IMPLEMENTATION_PLAN.md` - 勾选阶段任务
- `TODO.md` - 更新任务状态
- `MULTI_AGENT_PLAN.md` - 更新协作状态(如适用)
**步骤3.3:创建阶段提交**
```bash
git add .
git commit -m "feat([feature-name]): Complete Phase [N] - [Phase Name]
Phase [N] Summary:
- ✅ Task 1 completed
- ✅ Task 2 completed
- ✅ Task 3 completed
Tests: [X] passing, coverage [Y%]
Phase: [N] of [Total]
Next: Phase [N+1]"Phase 4: Final Integration & Handoff
阶段4:最终集成与交付
After all phases complete:
Step 4.1: Integration Testing
bash
undefined完成所有阶段后:
步骤4.1:集成测试
bash
undefinedMerge feature branch to develop
Merge feature branch to develop
git checkout develop
git merge --no-ff feature/[feature-name]
git checkout develop
git merge --no-ff feature/[feature-name]
Run full test suite including integration tests
Run full test suite including integration tests
npm run test:integration
npm run test:integration
Verify all tests pass
Verify all tests pass
**Step 4.2: Create Pull Request**
```bash
gh pr create \
--title "feat: [Feature Name]" \
--body-file PR_DESCRIPTION.md \
--label "ready-for-review"PR Description Template:
markdown
undefined
**步骤4.2:创建拉取请求(PR)**
```bash
gh pr create \
--title "feat: [Feature Name]" \
--body-file PR_DESCRIPTION.md \
--label "ready-for-review"PR描述模板:
markdown
undefinedFeature: [Name]
Feature: [Name]
Summary
Summary
[1-2 sentence description]
[1-2 sentence description]
Implementation Details
Implementation Details
- Component 1: [Description]
- Component 2: [Description]
- Component 1: [Description]
- Component 2: [Description]
Test Coverage
Test Coverage
- Unit tests: [X] tests, [Y%] coverage
- Integration tests: [Z] scenarios
- Unit tests: [X] tests, [Y%] coverage
- Integration tests: [Z] scenarios
Breaking Changes
Breaking Changes
- [None / List breaking changes]
- [None / List breaking changes]
Migration Guide
Migration Guide
[If breaking changes, provide migration steps]
[If breaking changes, provide migration steps]
Checklist
Checklist
- All tests passing
- Coverage >= 80%
- Linting passes
- Documentation updated
- No known bugs
**Step 4.3: Handoff to Validator**
If using multi-agent workflow, create handoff document:
```markdown
---
TO: Validator Agent (or use validator-role-skill)
FEATURE: [Feature Name]
PR: #[PR number]
IMPLEMENTATION_PLAN: IMPLEMENTATION_PLAN.md
TEST_COVERAGE: [X%]
IMPLEMENTATION_NOTES:
- [Key implementation detail 1]
- [Any concerns or trade-offs]
- [Areas needing special attention]
VALIDATION_REQUESTS:
- [ ] Unit test review
- [ ] Integration test verification
- [ ] Code quality assessment
- [ ] Security review (if handling sensitive data)
---- All tests passing
- Coverage >= 80%
- Linting passes
- Documentation updated
- No known bugs
**步骤4.3:交付给Validator**
如果使用多Agent工作流,创建交付文档:
```markdown
---
TO: Validator Agent (or use validator-role-skill)
FEATURE: [Feature Name]
PR: #[PR number]
IMPLEMENTATION_PLAN: IMPLEMENTATION_PLAN.md
TEST_COVERAGE: [X%]
IMPLEMENTATION_NOTES:
- [Key implementation detail 1]
- [Any concerns or trade-offs]
- [Areas needing special attention]
VALIDATION_REQUESTS:
- [ ] Unit test review
- [ ] Integration test verification
- [ ] Code quality assessment
- [ ] Security review (if handling sensitive data)
---Specialized Workflows
特殊工作流
Workflow A: Bug Fix Implementation
工作流A:Bug修复实现
When to Use: Fixing defects in existing code
Step A.1: Bug Analysis
markdown
undefined适用场景:修复现有代码中的缺陷
步骤A.1:Bug分析
markdown
undefinedBug Fix Plan: [Bug ID]
Bug Fix Plan: [Bug ID]
Problem Description
Problem Description
[What is broken, symptoms, reproduction steps]
[What is broken, symptoms, reproduction steps]
Root Cause Analysis
Root Cause Analysis
[Why it's broken - technical explanation]
[Why it's broken - technical explanation]
Proposed Solution
Proposed Solution
[How to fix it - specific approach]
[How to fix it - specific approach]
Affected Components
Affected Components
[List files/modules requiring changes]
[List files/modules requiring changes]
Regression Risk
Regression Risk
[What could potentially break]
[What could potentially break]
Testing Strategy
Testing Strategy
[How to verify fix + prevent regression]
**Step A.2: Test-Driven Fix**
1. **Write failing test that reproduces bug**:
```javascript
test('Bug #123: division by zero should throw error', () => {
const calculator = new Calculator();
expect(() => calculator.divide(10, 0))
.toThrow('Cannot divide by zero');
});- Implement minimal fix:
javascript
divide(a, b) {
if (b === 0) {
throw new Error('Cannot divide by zero');
}
return a / b;
}- Verify test passes
- Run full test suite (check for regressions)
- Commit with "fix:" prefix
Step A.3: Verification
bash
undefined[How to verify fix + prevent regression]
**步骤A.2:测试驱动的修复**
1. **编写可复现Bug的失败测试用例**:
```javascript
test('Bug #123: division by zero should throw error', () => {
const calculator = new Calculator();
expect(() => calculator.divide(10, 0))
.toThrow('Cannot divide by zero');
});- 实现最小化修复代码:
javascript
divide(a, b) {
if (b === 0) {
throw new Error('Cannot divide by zero');
}
return a / b;
}- 验证测试通过
- 运行完整测试套件(检查是否引入回归问题)
- 以"fix:"前缀提交代码
步骤A.3:验证
bash
undefinedRun affected component tests
Run affected component tests
npm test -- --testPathPattern=[component]
npm test -- --testPathPattern=[component]
Run full suite
Run full suite
npm test
npm test
Manual verification if UI/UX involved
Manual verification if UI/UX involved
[Steps to manually verify fix]
---[Steps to manually verify fix]
---Workflow B: Refactoring Implementation
工作流B:重构实现
When to Use: Improving code structure without changing behavior
Step B.1: Refactoring Justification
markdown
undefined适用场景:在不改变功能的前提下优化代码结构
步骤B.1:重构合理性说明
markdown
undefinedRefactoring Proposal: [Component Name]
Refactoring Proposal: [Component Name]
Current Problems
Current Problems
- [Problem 1: e.g., Code duplication]
- [Problem 2: e.g., Poor naming]
- [Problem 3: e.g., High complexity]
- [Problem 1: e.g., Code duplication]
- [Problem 2: e.g., Poor naming]
- [Problem 3: e.g., High complexity]
Proposed Improvements
Proposed Improvements
- [Improvement 1: Extract common logic]
- [Improvement 2: Rename variables for clarity]
- [Improvement 3: Split large function]
- [Improvement 1: Extract common logic]
- [Improvement 2: Rename variables for clarity]
- [Improvement 3: Split large function]
Risk Assessment
Risk Assessment
- Breaking changes: [Yes/No]
- Current test coverage: [X%]
- Effort estimate: [Hours]
- Architect approval required: [Yes/No]
**Step B.2: Safety-First Refactoring**
1. **Ensure comprehensive test coverage FIRST**
- If coverage < 80%, write tests before refactoring
- Tests act as safety net
2. **Make incremental changes**
- One refactoring at a time
- Commit after each logical change
- Run tests after EVERY change
3. **Never break public APIs** without version bump
- Internal refactoring OK
- Public API changes require coordination
4. **Document breaking changes** clearly
- Update CHANGELOG.md
- Provide migration guide
- Notify stakeholders
---- Breaking changes: [Yes/No]
- Current test coverage: [X%]
- Effort estimate: [Hours]
- Architect approval required: [Yes/No]
**步骤B.2:安全优先的重构**
1. **首先确保全面的测试覆盖**
- 如果覆盖率低于80%,先编写测试用例再进行重构
- 测试用例作为重构的安全保障
2. **进行增量式修改**
- 一次只进行一项重构
- 完成每个逻辑单元的修改后提交代码
- 每次修改后都运行测试
3. **无版本升级时绝不破坏公共API**
- 内部重构不受限制
- 修改公共API需要协调相关人员
4. **清晰记录破坏性变更**
- 更新CHANGELOG.md
- 提供迁移指南
- 通知相关利益方
---Code Quality Standards
代码质量标准
File-Level Requirements
文件级要求
Every file must have:
- File-level docstring/comment explaining purpose
- Appropriate imports/dependencies
- Consistent formatting (via auto-formatter)
- Error handling for failure modes
- Input validation where applicable
每个文件必须包含:
- 说明文件用途的文件级注释/文档字符串
- 合理的导入/依赖
- 一致的代码格式(通过自动格式化工具)
- 针对故障场景的错误处理
- 适用场景下的输入验证
Function-Level Requirements
函数级要求
Every function must have:
- Clear, descriptive name (verb for actions)
- Docstring/comment (purpose, params, returns)
- Type hints/annotations (if language supports)
- Single responsibility principle
- Unit test coverage
Example (TypeScript):
typescript
/**
* Validates user email format and domain
* @param email - Email address to validate
* @returns true if valid, false otherwise
* @throws ValidationError if email is null/undefined
*/
function validateUserEmail(email: string): boolean {
if (!email) {
throw new ValidationError('Email is required');
}
return /^[^\s@]+@[^\s@]+\.[^\s@]+$/.test(email);
}每个函数必须包含:
- 清晰、描述性的名称(动作使用动词)
- 文档字符串/注释(说明用途、参数、返回值)
- 类型提示/注解(如果语言支持)
- 单一职责原则
- 单元测试覆盖
示例(TypeScript):
typescript
/**
* Validates user email format and domain
* @param email - Email address to validate
* @returns true if valid, false otherwise
* @throws ValidationError if email is null/undefined
*/
function validateUserEmail(email: string): boolean {
if (!email) {
throw new ValidationError('Email is required');
}
return /^[^\s@]+@[^\s@]+\.[^\s@]+$/.test(email);
}Class-Level Requirements
类级要求
Every class must have:
- Class-level docstring
- Well-defined public interface
- Private methods clearly marked
- Constructor documentation
- Test coverage of public methods
每个类必须包含:
- 类级文档字符串
- 定义清晰的公共接口
- 明确标记私有方法
- 构造函数文档
- 公共方法的测试覆盖
Commit Message Standards
提交信息标准
Use Conventional Commits format:
<type>(<scope>): <subject>
<body>
<footer>Types:
- : New feature
feat - : Bug fix
fix - : Documentation only
docs - : Formatting, missing semi-colons, etc.
style - : Code restructuring
refactor - : Adding/updating tests
test - : Maintenance tasks
chore
Example:
feat(auth): add JWT token refresh mechanism
- Implement refresh token endpoint
- Add token expiration validation
- Update authentication middleware
- Add unit tests for refresh flow
Tests: 45 passing, coverage 92%
Closes: #123采用Conventional Commits格式:
<type>(<scope>): <subject>
<body>
<footer>类型:
- : 新功能
feat - : Bug修复
fix - : 仅更新文档
docs - : 格式调整,如缺少分号等
style - : 代码重构
refactor - : 添加/更新测试
test - : 维护任务
chore
示例:
feat(auth): add JWT token refresh mechanism
- Implement refresh token endpoint
- Add token expiration validation
- Update authentication middleware
- Add unit tests for refresh flow
Tests: 45 passing, coverage 92%
Closes: #123Quality Assurance Checklists
质量保障检查清单
Before Each Commit
每次提交前
- Code compiles/runs without errors
- All tests pass
- Linter passes with zero errors
- Code formatted with project formatter
- No commented-out code blocks
- No debug logging statements
- Commit message follows convention
- 代码可编译/运行无错误
- 所有测试通过
- 代码检查工具检测无错误
- 代码已通过项目格式化工具处理
- 无注释掉的代码块
- 无调试日志语句
- 提交信息符合规范
Before Phase Completion
阶段完成前
- Phase tasks all complete
- Test coverage meets threshold (80%+)
- Documentation updated
- IMPLEMENTATION_PLAN.md updated
- No known bugs or issues
- Performance acceptable
- 阶段所有任务已完成
- 测试覆盖率达到阈值(80%+)
- 文档已更新
- IMPLEMENTATION_PLAN.md已更新
- 无已知Bug或问题
- 性能符合要求
Before PR Creation
创建PR前
- All phases complete
- Integration tests pass
- No merge conflicts with target branch
- PR description complete and accurate
- Breaking changes documented
- Migration guide provided (if needed)
- 所有阶段已完成
- 集成测试通过
- 与目标分支无合并冲突
- PR描述完整准确
- 破坏性变更已记录
- 已提供迁移指南(如需要)
Collaboration Patterns
协作模式
With Architect (or architect-role-skill)
与Architect(或architect-role-skill)协作
- Request clarification on unclear specifications
- Propose alternative implementations when issues discovered
- Report architectural concerns immediately
- Never deviate from architecture without approval
- 对不明确的规范请求澄清
- 发现问题时提出替代实现方案
- 立即反馈架构相关问题
- 未获批准绝不偏离架构设计
With Validator (or validator-role-skill)
与Validator(或validator-role-skill)协作
- Provide comprehensive test instructions
- Document known limitations or edge cases
- Request specific security reviews when handling sensitive data
- Respond promptly to code review feedback
- 提供全面的测试说明
- 记录已知限制或边缘情况
- 处理敏感数据时请求特定安全评审
- 及时响应代码评审反馈
With Scribe (or scribe-role-skill)
与Scribe(或scribe-role-skill)协作
- Update inline documentation as code changes
- Flag complex algorithms needing detailed explanation
- Provide API usage examples
- Document breaking changes
- 代码变更时同步更新内联文档
- 标记需要详细说明的复杂算法
- 提供API使用示例
- 记录破坏性变更
With DevOps (or devops-role-skill)
与DevOps(或devops-role-skill)协作
- Communicate new dependencies or environment requirements
- Provide database migration scripts
- Document configuration changes
- Alert to performance-critical changes
- 告知新的依赖或环境要求
- 提供数据库迁移脚本
- 记录配置变更
- 提醒性能关键型变更
Error Handling Protocols
错误处理协议
When Stuck on Implementation
实现遇到阻碍时
- Document the problem in IMPLEMENTATION_PLAN.md
- Research similar patterns in codebase (use grep/search)
- Consult external documentation
- Use researcher-role-skill if needed for deep investigation
- Escalate to architect-role-skill if architectural change needed
- 在IMPLEMENTATION_PLAN.md中记录问题
- 在代码库中搜索类似模式(使用grep/搜索工具)
- 查询外部文档
- 如需深度调研可使用researcher-role-skill
- 如需架构变更则升级至architect-role-skill
When Tests Fail
测试失败时
- Analyze test failure output
- Debug with focused console logging
- Isolate failing component
- Fix or update test as appropriate
- Never skip or disable tests to make them pass
- 分析测试失败输出
- 通过聚焦式控制台日志调试
- 隔离失败组件
- 修复或更新测试用例
- 绝不通过跳过或禁用测试来让测试通过
When Dependencies Conflict
依赖冲突时
- Document the conflict
- Research resolution in package documentation
- Test resolution in isolated environment
- Update dependency management files
- Notify DevOps (or use devops-role-skill) of environment changes
- 记录冲突情况
- 在包文档中查询解决方案
- 在隔离环境中测试解决方案
- 更新依赖管理文件
- 告知DevOps(或使用devops-role-skill)环境变更
Performance Considerations
性能考量
Code Efficiency Guidelines
代码效率指南
- Optimize only when profiling shows bottleneck
- Prefer clarity over premature optimization
- Use appropriate data structures (arrays vs objects vs sets)
- Avoid N+1 queries (use eager loading, joins)
- Cache expensive computations
- Consider pagination for large datasets
- 仅在性能分析显示存在瓶颈时才进行优化
- 优先保证代码清晰,避免过早优化
- 使用合适的数据结构(数组、对象、集合)
- 避免N+1查询(使用预加载、关联查询)
- 缓存昂贵的计算结果
- 针对大型数据集考虑分页
Resource Management
资源管理
- Close file handles and connections
- Manage memory in long-running processes
- Use connection pooling for databases
- Implement timeouts for external API calls
- Log resource usage in development
- 关闭文件句柄与连接
- 在长期运行的进程中管理内存
- 对数据库使用连接池
- 为外部API调用实现超时机制
- 在开发环境中记录资源使用情况
Security Implementation Standards
安全实现标准
Input Validation
输入验证
- Validate ALL user input
- Sanitize before database queries
- Use parameterized queries (NEVER string concatenation)
- Validate file uploads (type, size, content)
- Implement rate limiting where appropriate
Example (SQL Injection Prevention):
javascript
// ❌ WRONG - Vulnerable to SQL injection
const query = `SELECT * FROM users WHERE email = '${userEmail}'`;
// ✅ CORRECT - Parameterized query
const query = 'SELECT * FROM users WHERE email = ?';
db.execute(query, [userEmail]);- 验证所有用户输入
- 存入数据库前进行清理
- 使用参数化查询(绝不使用字符串拼接)
- 验证文件上传(类型、大小、内容)
- 适用场景下实现速率限制
示例(SQL注入防护):
javascript
// ❌ 错误写法 - 易受SQL注入攻击
const query = `SELECT * FROM users WHERE email = '${userEmail}'`;
// ✅ 正确写法 - 参数化查询
const query = 'SELECT * FROM users WHERE email = ?';
db.execute(query, [userEmail]);Authentication & Authorization
认证与授权
- Never store passwords in plain text
- Use established libraries for crypto operations
- Validate authorization on every request
- Implement proper session management
- Log authentication events
- 绝不明文存储密码
- 使用成熟的加密库
- 对每个请求验证授权
- 实现合理的会话管理
- 记录认证事件
Data Protection
数据保护
- Encrypt sensitive data at rest
- Use HTTPS for data in transit
- Redact sensitive info from logs
- Implement proper access controls
- Follow principle of least privilege
- 静态敏感数据加密
- 传输数据使用HTTPS
- 日志中脱敏敏感信息
- 实现合理的访问控制
- 遵循最小权限原则
Examples
示例
Example 1: Simple Feature Implementation
示例1:简单功能实现
Task: Add user registration endpoint
markdown
undefined任务:添加用户注册接口
markdown
undefinedPhase 1: Foundation
Phase 1: Foundation
- Write test for user model validation
- Create user model with email, password fields
- Add email format validation
- Add password strength validation
- Write test for user model validation
- Create user model with email, password fields
- Add email format validation
- Add password strength validation
Phase 2: Business Logic
Phase 2: Business Logic
- Write test for registration service
- Implement registration service
- Add duplicate email check
- Hash password before storage
- Write test for registration service
- Implement registration service
- Add duplicate email check
- Hash password before storage
Phase 3: API Integration
Phase 3: API Integration
- Write integration test for /register endpoint
- Create POST /register endpoint
- Add input validation middleware
- Add error handling
Result: Feature complete in 3 phases with 95% test coverage
undefined- Write integration test for /register endpoint
- Create POST /register endpoint
- Add input validation middleware
- Add error handling
Result: Feature complete in 3 phases with 95% test coverage
undefinedExample 2: Bug Fix
示例2:Bug修复
Task: Fix user login timeout issue
markdown
undefined任务:修复用户登录超时问题
markdown
undefinedBug Analysis
Bug Analysis
- Problem: Login hangs after 30 seconds
- Root cause: Database query missing index on email column
- Solution: Add database index
- Problem: Login hangs after 30 seconds
- Root cause: Database query missing index on email column
- Solution: Add database index
Implementation
Implementation
- Write test that times login query
- Add migration script for index
- Run migration
- Verify test passes
- Measure performance improvement (30s → 50ms)
undefined- Write test that times login query
- Add migration script for index
- Run migration
- Verify test passes
- Measure performance improvement (30s → 50ms)
undefinedExample 3: Refactoring
示例3:重构
Task: Extract duplicate validation logic
markdown
undefined任务:提取重复的验证逻辑
markdown
undefinedCurrent Problem
Current Problem
- Email validation duplicated in 5 files
- Password validation duplicated in 3 files
- Email validation duplicated in 5 files
- Password validation duplicated in 3 files
Solution
Solution
- Ensure all 5 files have tests (add if missing)
- Extract common validation to utils/validators.js
- Update imports in all 5 files
- Run tests after each file update
- Remove old validation code
- Final test run - all pass
Result: Code duplication eliminated, tests still pass
---- Ensure all 5 files have tests (add if missing)
- Extract common validation to utils/validators.js
- Update imports in all 5 files
- Run tests after each file update
- Remove old validation code
- Final test run - all pass
Result: Code duplication eliminated, tests still pass
---Resources
资源
Templates
模板
- - Implementation plan template
resources/IMPLEMENTATION_PLAN_template.md - - Commit message examples
resources/commit_message_guide.md - - TDD cycle reference
resources/tdd_workflow.md
- - 实现计划模板
resources/IMPLEMENTATION_PLAN_template.md - - 提交信息示例
resources/commit_message_guide.md - - TDD循环参考
resources/tdd_workflow.md
Scripts
脚本
- - Test execution script
scripts/run_tests.sh - - Pre-commit validation
scripts/validate_commit.py - - Phase completion validator
scripts/phase_checker.sh
- - 测试执行脚本
scripts/run_tests.sh - - 提交前验证脚本
scripts/validate_commit.py - - 阶段完成验证脚本
scripts/phase_checker.sh
References
参考
- Agent Skills vs. Multi-Agent
- Test-Driven Development Skill (if available)
- Git Worktrees Skill
Version: 1.0.0
Last Updated: December 12, 2025
Status: ✅ Active
Maintained By: Claude Command and Control Project
- Agent Skills vs. Multi-Agent
- Test-Driven Development Skill (if available)
- Git Worktrees Skill
版本: 1.0.0
最后更新: 2025年12月12日
状态: ✅ 活跃
维护方: Claude Command and Control Project