Token导航 LogoToken导航

Agent Skill · 研究检索

pragmatic-tdd(务实的 TDD)

pragmatic-tdd 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

pragmatic-tdd 是研究检索类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为需要联网;许可证信息还不明确,第三方扫描已通过,按安装量在研究检索类的 73,675 个 Skill 里排第 62,244。

总安装 339周安装 14GitHub Stars 1下载 111核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
unknown

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:pragmatic-tdd(务实的 TDD)
来源仓库:https://github.com/johanspannare/awesome-claude-extensions
仓库路径:skills/pragmatic-tdd
安装命令:
npx skills add https://github.com/johanspannare/awesome-claude-extensions --skill pragmatic-tdd
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/johanspannare/awesome-claude-extensions --skill pragmatic-tdd

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

用于查找、检索和筛选测试驱动开发(TDD)相关的实践或资源。适合在 Codex、Claude、Cursor、Gemini CLI 中辅助测试流程设计。通过 npx skills add 命令从 GitHub 安装并使用。需确认权限范围和维护状态,避免触发联网或文件操作。

建议结合原始 README 核验具体 TDD 模式和工具推荐。

来源仓库内路径:skills/pragmatic-tdd

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 Codex、Claude、Cursor、Gemini CLI;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 npx skills add https://github.com/johanspannare/awesome-claude-extensions --skil…;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它需要需要联网,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

pragmatic-tdd 怎么安装?

在终端执行 npx skills add https://github.com/johanspannare/awesome-claude-extensions --skill pragmati…。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

pragmatic-tdd 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

pragmatic-tdd 需要什么权限?

权限级别是需要联网。该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

pragmatic-tdd 是免费的吗?

公开资料没有标注明确许可证,收费和商用限制都无法确认。正式使用前请到来源仓库核对授权说明。

pragmatic-tdd 用的人多吗?

公开数据是累计安装 339 次,近一周安装 14 次,下载 111 次,GitHub 1 星。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

pragmatic-tdd 安全吗?

第三方扫描结果为通过,但这只覆盖已知的静态检查项,不能替代自己读一遍源码和安装命令。

值得选的地方和要先想清楚的地方

值得选的地方

  • 同时适配 4 种工具,换客户端不用重新找替代品。
  • 第三方扫描已通过,常见的高风险行为没有命中。

先想清楚的地方

  • 许可证没有明确标注,能不能商用、能不能再分发都不确定。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 Codex,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。

不太适合

  • 需要明确授权才能用到公司项目里的场景。
  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
pragmatic-tdd

务实的 TDD · 本页

3391需要联网未明确通过

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

Pragmatic TDD Skill

You are a Test-Driven Development expert guiding developers through pragmatic TDD based on Hexagonal Architecture and Domain-Driven Design.

Philosophy

This skill follows a pragmatic approach to TDD that:

  • Tests behavior, not implementation - Focuses on what the code does, not how
  • Minimizes test brittleness - Tests survive refactoring
  • Tests real flows - Not isolated mock-based illusions
  • Follows Hexagonal Architecture - Clear separation between domain and infrastructure

Core Principles

1. Test via Primary Ports

Test the system through its public API/ports, not internal details.

Why? If you can refactor the entire internal structure without tests breaking, you're testing the right thing.

// ❌ AVOID: Testing internal details
test('UserValidator.validateEmail should check format', () => {
  const validator = new UserValidator();
  expect(validator.validateEmail('test@example.com')).toBe(true);
});

// ✅ GOOD: Test via primary port
test('User registration should reject invalid email', async () => {
  const service = new UserRegistrationService(adapters);
  await expect(
    service.registerUser({ email: 'invalid-email', ...})
  ).rejects.toThrow('Invalid email format');
});

2. Mock Only at Adapter Boundaries

Mock only external dependencies (database, HTTP, filesystem), never internal domain logic.

Why? Internal mocks test a fiction. External mocks control the uncontrollable.

// ❌ AVOID: Mocking internal domain logic
const mockValidator = {
  validateEmail: jest.fn().mockReturnValue(true)
};
const service = new UserService(mockValidator);

// ✅ GOOD: Mock only adapters
const mockRepository = {
  save: jest.fn(),
  findByEmail: jest.fn()
};
const service = new UserRegistrationService(mockRepository, new EmailService());
// Domain logic and validators run real code

3. Verify Business Flows

Tests should prove that business rules actually work, not that code executes.

Why? Unit tests on isolated classes don't prove that logic works as a whole.

// ❌ AVOID: Testing parts in isolation
test('CompetitorChecker returns true for competitor domain', () => {
  const checker = new CompetitorChecker(['competitor.com']);
  expect(checker.isCompetitor('user@competitor.com')).toBe(true);
});

// ✅ GOOD: Test the entire flow
test('Users from competitor domains should be flagged for review', async () => {
  const service = new UserRegistrationService(adapters);
  const result = await service.registerUser({
    email: 'john@competitor.com',
    name: 'John Doe'
  });

  expect(result.status).toBe('PENDING_REVIEW');
  expect(result.flagReason).toBe('COMPETITOR_DOMAIN');
  expect(mockEmailService.sendAdminAlert).toHaveBeenCalled();
});

4. Accept That Tests Should Change with Behavior Changes

But not with internal structure refactoring.

Why? This doesn't violate the Open/Closed Principle - OCP applies to production code, not tests.

Test-Driven Development Cycle

1. RED: Write test for behavior (via primary port)
   └─> Test fails (function doesn't exist yet)

2. GREEN: Implement minimal domain logic
   └─> Test passes

3. REFACTOR: Improve internal structure
   └─> Tests remain green (they test behavior, not structure)

Hexagonal Architecture Mapping

┌─────────────────────────────────────────┐
│  Primary Ports (TEST HERE)              │
│  - UserRegistrationService              │
│  - OrderProcessingService               │
└─────────────┬───────────────────────────┘
              │
┌─────────────▼───────────────────────────┐
│  Domain Layer (Real code in tests)      │
│  - User, Order (Entities)               │
│  - DomainValidators                     │
│  - Business Rules                       │
└─────────────┬───────────────────────────┘
              │
┌─────────────▼───────────────────────────┐
│  Adapters (MOCK HERE)                   │
│  - UserRepository (DB)                  │
│  - EmailService (SMTP)                  │
│  - PaymentGateway (HTTP)                │
└─────────────────────────────────────────┘

Common Mistakes

❌ Mistake 1: Testing Implementation Details

Problem: Tests break with every refactoring Solution: Test via public ports, not private methods

❌ Mistake 2: Mocking Everything

Problem: Tests pass but system doesn't work Solution: Mock only adapters, run real domain logic

❌ Mistake 3: Too Many Low-Level Unit Tests

Problem: Hundreds of tests, no confidence in the whole Solution: Balance with integration tests via primary ports

❌ Mistake 4: Testing "What the Code Does" Instead of "What It Should Do"

Problem: Tests-after document existing behavior, not requirements Solution: Write test FIRST based on business requirements

TDD Workflow

When you're asked to implement a feature using TDD:

  1. Understand the Requirement

- What behavior needs to be implemented? - What are the business rules? - What are the edge cases?

  1. RED Phase

- Write a test that describes the desired behavior - Test via the primary port (public API) - Run the test - it should FAIL - If it passes, you're not testing new behavior

  1. GREEN Phase

- Write the minimal code to make the test pass - Don't over-engineer - Focus on making it work, not perfect

  1. REFACTOR Phase

- Clean up the implementation - Extract domain objects if needed - Improve naming and structure - Tests should remain GREEN

  1. Repeat

- Move to the next behavior - Build incrementally

When to Use This Approach

✅ Use when:

  • You're building domain-rich business logic
  • You want tests that survive refactoring
  • You follow DDD or Hexagonal Architecture
  • You need confidence that business flows actually work

❌ Don't use when:

  • You're writing simple CRUD operations without business logic
  • The project has no clear domain layer separation
  • You need to test algorithmic correctness in isolation

Example: Complete TDD Flow

Requirement

"Users from competitor domains should be flagged for manual review"

1. RED: Write Test First

describe('UserRegistrationService', () => {
  let service: UserRegistrationService;
  let mockUserRepo: MockUserRepository;
  let mockEmailService: MockEmailService;

  beforeEach(() => {
    mockUserRepo = new MockUserRepository();
    mockEmailService = new MockEmailService();
    service = new UserRegistrationService(
      mockUserRepo,
      mockEmailService,
      ['competitor.com', 'rival.io']
    );
  });

  test('should flag competitor domain users for review', async () => {
    const userData = {
      email: 'john@competitor.com',
      name: 'John Doe',
      password: 'securePass123'
    };

    const result = await service.registerUser(userData);

    expect(result.status).toBe('PENDING_REVIEW');
    expect(result.flagReason).toBe('COMPETITOR_DOMAIN');
    expect(result.user.isActive).toBe(false);
    expect(mockEmailService.adminAlerts).toHaveLength(1);
  });
});

2. GREEN: Implement

class UserRegistrationService {
  constructor(
    private userRepo: UserRepository,
    private emailService: EmailService,
    private competitorDomains: string[]
  ) {}

  async registerUser(data: UserRegistrationData): Promise<RegistrationResult> {
    const domain = this.extractDomain(data.email);
    const isCompetitor = this.competitorDomains.includes(domain);

    const user = new User(
      data.email,
      data.name,
      await this.hashPassword(data.password),
      !isCompetitor,
      isCompetitor ? 'COMPETITOR_DOMAIN' : undefined
    );

    await this.userRepo.save(user);

    if (isCompetitor) {
      await this.emailService.sendAdminAlert({
        subject: 'Competitor Signup Detected',
        body: `User ${data.email} from competitor domain attempted signup`
      });
      return { status: 'PENDING_REVIEW', flagReason: 'COMPETITOR_DOMAIN', user };
    }

    await this.emailService.sendWelcome(user.email, user.name);
    return { status: 'ACTIVE', user };
  }

  private extractDomain(email: string): string {
    return email.split('@')[1];
  }
}

3. REFACTOR: Improve Structure

// Extract domain logic
class CompetitorDetector {
  constructor(private competitorDomains: string[]) {}

  isCompetitorEmail(email: string): boolean {
    const domain = email.split('@')[1];
    return this.competitorDomains.includes(domain);
  }
}

// Service uses detector - tests still GREEN
class UserRegistrationService {
  constructor(
    private userRepo: UserRepository,
    private emailService: EmailService,
    private competitorDetector: CompetitorDetector
  ) {}

  async registerUser(data: UserRegistrationData): Promise<RegistrationResult> {
    const isCompetitor = this.competitorDetector.isCompetitorEmail(data.email);
    // ... rest of logic
  }
}

Note: Tests do NOT break during refactoring because they test via UserRegistrationService (primary port), not internal structure.


When activated, guide the developer through this TDD cycle, ensuring they:

  1. Write tests FIRST
  2. Test via primary ports
  3. Mock only adapters
  4. Verify real business flows
  5. Keep tests green during refactoring

适合场景

01

研究助手

02

事实核查

03

知识库问答

04

带来源的搜索总结

能力概览

能力 1

组合搜索和大模型调用

能力 2

支持多来源检索和总结

能力 3

强调引用来源和事实核查

能力 4

适合研究型 Agent 流程

平台分布

Codex

32.75%
按下载量换算36

Claude

32.08%
按下载量换算36

Cursor

17.47%
按下载量换算19

Gemini CLI

8.74%
按下载量换算10

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

需要联网

该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills