Token导航 LogoToken导航

Agent Skill · 研究检索

testing-trophy(测试奖杯)

用于辅助测试设计、自动化测试、用例整理和回归验证。它适合让 Agent 编写单元测试、端到端测试、测试计划或根据失败日志定位问题。使用时需要确认项目测试框架、运行命令和夹具数据,避免为了通过测试而改坏真实逻辑;涉及浏览器或外部服务时,应区分本地模拟、测试环境和生产环境。

testing-trophy 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,权限为操作浏览器;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 71,089。

总安装 196周安装 8GitHub Stars 2下载 63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:testing-trophy(测试奖杯)
来源仓库:https://github.com/jonmumm/skills
仓库路径:skills/testing-trophy
安装命令:
npx skills add https://github.com/jonmumm/skills --skill testing-trophy
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/jonmumm/skills --skill testing-trophy

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。

用于辅助测试设计、自动化测试、用例整理和回归验证。适合编写单元测试、端到端测试、测试计划或根据失败日志定位问题。使用时需确认项目测试框架、运行命令和夹具数据,避免为通过测试而改坏真实逻辑。涉及浏览器或外部服务时,应区分本地模拟、测试环境和生产环境。

安装方式:github,支持 Codex、Claude、Cursor、Gemini CLI。

来源仓库内路径:skills/testing-trophy

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/jonmumm/skills --skill testing-trophy;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

    它需要操作浏览器,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

testing-trophy 怎么安装?

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

testing-trophy 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。

testing-trophy 需要什么权限?

权限级别是操作浏览器。该 Skill 可能涉及浏览器控制能力,使用时可能读取或操作网页内容,需要在受控环境中确认权限边界。

testing-trophy 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。

testing-trophy 用的人多吗?

公开数据是累计安装 196 次,近一周安装 8 次,下载 63 次,GitHub 2 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。

testing-trophy 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
testing-trophy

测试奖杯 · 本页

1962操作浏览器未明确通过

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

Testing Trophy

You follow Kent C. Dodds' Testing Trophy philosophy. The trophy shape means:

        🏆
      E2E Tests        ← Few, slow, high confidence
    Integration Tests   ← MOST tests live here
   Unit Tests           ← Some, for complex logic only
 Static Analysis        ← TypeScript, ESLint, tsc

The key insight: Write tests that give you confidence your app works for users. Integration tests hit the sweet spot — they test real behavior through real boundaries without the brittleness of E2E or the false confidence of isolated unit mocks.

Core Principles

  1. Test behavior, not implementation. Assert on what the user sees or what the API returns — not internal state.
  2. Don't mock what you own. Mock external services (Stripe, Expo Push), not your own modules.
  3. One test > ten mocks. A single integration test through the real stack catches more bugs than ten unit tests with mocked dependencies.
  4. The more your tests resemble how your software is used, the more confidence they give you.
  5. Coverage is a side effect, not a goal. High coverage with bad tests is worse than moderate coverage with good tests.

When to Use Each Level

LevelWhenExample
StaticAlways — it's freetsc --noEmit, ESLint, Zod schemas
UnitComplex pure logic, algorithms, state machinesScoring functions, parsers, reducers
IntegrationDefault for everything elseAPI routes with real D1, React components with real state
E2ECritical user journeys, smoke tests"User can create a game and invite players"

React: Storybook + Play Functions as Integration Tests

Storybook play functions ARE integration tests. They render real components, interact with real DOM, and assert on real behavior — in a real browser.

Pattern: Component Integration via Play Functions

// CastButton.stories.tsx
import type { Meta, StoryObj } from '@storybook/react';
import { within, userEvent, expect } from '@storybook/test';
import { CastButton } from './CastButton';
import { CastProvider } from '@open-game-system/cast-kit-react';

const meta: Meta<typeof CastButton> = {
  component: CastButton,
  decorators: [
    (Story) => (
      <CastProvider>
        <Story />
      </CastProvider>
    ),
  ],
};
export default meta;

type Story = StoryObj<typeof CastButton>;

// Story IS the test — play function exercises real behavior
export const ShowsCastAvailable: Story = {
  play: async ({ canvasElement }) => {
    const canvas = within(canvasElement);

    // Real component renders with real provider
    const button = await canvas.findByRole('button', { name: /cast/i });
    await expect(button).toBeVisible();

    // Interact like a user
    await userEvent.click(button);

    // Assert on what the user sees
    await expect(canvas.getByText(/available/i)).toBeVisible();
  },
};

export const ConnectedState: Story = {
  parameters: {
    // Mock the bridge state, not the component internals
    castState: {
      isAvailable: true,
      session: { status: 'connected', deviceName: 'Living Room TV' },
    },
  },
  play: async ({ canvasElement }) => {
    const canvas = within(canvasElement);
    await expect(canvas.getByText(/connected/i)).toBeVisible();
    await expect(canvas.getByText(/Living Room TV/i)).toBeVisible();
  },
};

Why This Beats Unit Tests

// ❌ BAD: Unit test with mocks — tests implementation, not behavior
test('CastButton calls dispatch with SHOW_CAST_PICKER', () => {
  const mockDispatch = vi.fn();
  vi.mock('~/hooks/useCastDispatch', () => ({ useCastDispatch: () => mockDispatch }));
  render(<CastButton />);
  fireEvent.click(screen.getByRole('button'));
  expect(mockDispatch).toHaveBeenCalledWith({ type: 'SHOW_CAST_PICKER' });
});

// ✅ GOOD: Integration test via Storybook play — tests real behavior
export const TapShowsPicker: Story = {
  play: async ({ canvasElement }) => {
    const canvas = within(canvasElement);
    await userEvent.click(canvas.getByRole('button'));
    // Assert on what appears, not what was called
    await expect(canvas.getByText(/select a device/i)).toBeVisible();
  },
};

Running Storybook Tests in CI

# Build Storybook, then run all play functions as tests
npx storybook build
npx test-storybook --ci

Reference: Use with /react-composable-components and /dont-use-use-effect skills.

Cloudflare Workers: vitest-pool-workers + Real D1

@cloudflare/vitest-pool-workers runs tests inside the real Workers runtime with real bindings. No mocks needed for D1, KV, R2.

Pattern: Full-Stack Integration Test

// test/integration/notifications.test.ts
import { env, SELF } from "cloudflare:test";
import { describe, it, expect, beforeEach } from "vitest";

describe("Notification Flow — Full Integration", () => {
  beforeEach(async () => {
    // Real D1 — clean between tests
    await env.DB.prepare("DELETE FROM devices").run();
  });

  it("registers device, sends notification end-to-end", async () => {
    // Step 1: Register device (real D1 insert)
    const registerRes = await SELF.fetch("https://api.test/api/v1/devices/register", {
      method: "POST",
      headers: { "Content-Type": "application/json" },
      body: JSON.stringify({
        ogsDeviceId: "device-1",
        platform: "ios",
        pushToken: "ExponentPushToken[xxx]",
      }),
    });
    expect(registerRes.status).toBe(200);
    const { deviceToken } = await registerRes.json();

    // Verify in real D1
    const device = await env.DB.prepare(
      "SELECT * FROM devices WHERE ogs_device_id = ?"
    ).bind("device-1").first();
    expect(device.push_token).toBe("ExponentPushToken[xxx]");

    // Step 2: Send notification (real auth + real D1 lookup)
    const sendRes = await SELF.fetch("https://api.test/api/v1/notifications/send", {
      method: "POST",
      headers: {
        "Content-Type": "application/json",
        Authorization: "Bearer test-api-key",
      },
      body: JSON.stringify({
        deviceToken,
        notification: { title: "Game Starting!", body: "Join now" },
      }),
    });

    // Push provider may fail in test env — that's fine
    // What matters: auth passed, D1 lookup worked, request reached the provider
    expect([200, 502]).toContain(sendRes.status);
  });
});

Setup: vitest.integration.config.mts

import { cloudflareTest } from "@cloudflare/vitest-pool-workers";
import { defineConfig } from "vitest/config";

export default defineConfig({
  plugins: [
    cloudflareTest({
      wrangler: { configPath: "./wrangler.toml" },
      miniflare: {
        bindings: {
          OGS_JWT_SECRET: "test-secret",
        },
      },
    }),
  ],
  test: {
    globals: true,
    include: ["test/integration/**/*.test.ts"],
    setupFiles: ["./test/integration/setup.ts"],
  },
});

Setup: Apply D1 Schema

// test/integration/setup.ts
import { env } from "cloudflare:test";

const schema = `
CREATE TABLE IF NOT EXISTS devices (...);
CREATE TABLE IF NOT EXISTS api_keys (...);
`;

for (const stmt of schema.split(";").filter(s => s.trim())) {
  await env.DB.prepare(stmt).run();
}

// Seed test data
await env.DB.prepare(
  "INSERT OR IGNORE INTO api_keys (key, game_id, game_name) VALUES (?, ?, ?)"
).bind("test-api-key", "test-game", "Test Game").run();

Why This Beats Mocked Unit Tests

// ❌ BAD: Mocked D1 — doesn't test real SQL, real bindings
const mockDB = { prepare: vi.fn(() => ({ bind: vi.fn(() => ({ first: vi.fn() })) })) };
const res = await app.request("/api/v1/devices/register", {}, { DB: mockDB });
expect(mockDB.prepare).toHaveBeenCalled(); // Tests mock wiring, not behavior

// ✅ GOOD: Real Workers runtime, real D1
const res = await SELF.fetch("https://api.test/api/v1/devices/register", { ... });
const row = await env.DB.prepare("SELECT * FROM devices WHERE ...").first();
expect(row.push_token).toBe("ExponentPushToken[xxx]"); // Tests real behavior

Reference: Use with /workers-integration-testing, /seam-tester, and /workers-best-practices skills.

Swift: XCTest UI Tests as Integration Tests

Swift's equivalent of Storybook play functions is XCTest UI Testing combined with SwiftUI Previews + Snapshot Tests.

Pattern: XCTest UI Integration Tests

// GameSetupUITests.swift
import XCTest

final class GameSetupUITests: XCTestCase {
    let app = XCUIApplication()

    override func setUpWithError() throws {
        continueAfterFailure = false
        app.launchArguments = ["--ui-testing"]
        app.launch()
    }

    func testCastButtonAppearsWhenDeviceAvailable() throws {
        // Navigate to game setup
        app.buttons["Create Game"].tap()

        // Cast button should appear (real UI, real state)
        let castButton = app.buttons["Cast to TV"]
        XCTAssertTrue(castButton.waitForExistence(timeout: 5))
        XCTAssertTrue(castButton.isEnabled)
    }

    func testCastButtonShowsDevicePicker() throws {
        app.buttons["Create Game"].tap()
        app.buttons["Cast to TV"].tap()

        // Device picker sheet should appear
        let picker = app.sheets["Select a device"]
        XCTAssertTrue(picker.waitForExistence(timeout: 3))
        XCTAssertTrue(picker.staticTexts["Living Room TV"].exists)
    }
}

Pattern: SwiftUI Preview + Snapshot Testing

// Using swift-snapshot-testing (pointfreeco)
import SnapshotTesting
import SwiftUI

final class CastButtonSnapshotTests: XCTestCase {
    func testCastButtonStates() {
        // Available state
        assertSnapshot(
            of: CastButton(state: .available(deviceCount: 2)),
            as: .image(layout: .fixed(width: 200, height: 44))
        )

        // Connected state
        assertSnapshot(
            of: CastButton(state: .connected(deviceName: "Living Room TV")),
            as: .image(layout: .fixed(width: 200, height: 44))
        )

        // Connecting state
        assertSnapshot(
            of: CastButton(state: .connecting),
            as: .image(layout: .fixed(width: 200, height: 44))
        )
    }
}

Testing Trophy for Swift

LevelToolWhat
StaticSwiftLint, Swift compilerType safety, conventions
UnitXCTestPure logic: scoring, parsing, state machines
IntegrationXCTest UI TestsReal app, real navigation, real state
E2EXCTest UI + real backendFull user journeys with real API

React Native (Expo) Equivalent: Detox

For React Native apps, Detox is the integration test layer:

// e2e/cast-flow.test.ts
import { by, device, element, expect, waitFor } from 'detox';

describe('Cast Flow', () => {
  beforeAll(async () => {
    await device.launchApp({ newInstance: true });
  });

  it('shows cast button when Chromecast is available', async () => {
    await element(by.text('Trivia Jam')).tap();
    await waitFor(element(by.id('castButton')))
      .toBeVisible()
      .withTimeout(10000);
  });

  it('opens device picker on cast button tap', async () => {
    await element(by.id('castButton')).tap();
    await expect(element(by.text('Select a device'))).toBeVisible();
  });
});

Anti-Patterns to Avoid

1. Testing Implementation Details

// ❌ Tests mock wiring, not behavior
expect(mockDispatch).toHaveBeenCalledWith({ type: 'START_CASTING' });

// ✅ Tests what the user sees
await expect(canvas.getByText('Connected to Living Room TV')).toBeVisible();

2. Excessive Mocking

// ❌ Everything is mocked — test proves nothing
vi.mock('./database');
vi.mock('./auth');
vi.mock('./notifications');
const result = await handler(mockReq);
expect(mockDB.insert).toHaveBeenCalled();

// ✅ Real stack, real assertions
const res = await SELF.fetch("https://api.test/...", { ... });
const row = await env.DB.prepare("SELECT ...").first();
expect(row).toBeTruthy();

3. Testing Library Internals

// ❌ Testing that React rendered correctly
expect(wrapper.find('CastButton').props().isAvailable).toBe(true);

// ✅ Testing that the button is visible to the user
await expect(screen.getByRole('button', { name: /cast/i })).toBeVisible();

4. Snapshot Abuse

// ❌ Giant snapshot that breaks on every CSS change
expect(component).toMatchSnapshot();

// ✅ Targeted snapshot of specific states
assertSnapshot(of: CastButton(state: .connected("TV")), as: .image(...));

Related Skills

  • /seam-tester — Integration tests at system boundaries
  • /workers-integration-testing — Cloudflare Workers with vitest-pool-workers
  • /mutation-testing — Verify test quality after writing integration tests
  • /tdd — Red-green-refactor with integration-first approach
  • /e2e-testing-patterns — Playwright/Cypress for the top of the trophy
  • /react-composable-components — Components that are easy to integration-test
  • /dont-use-use-effect — Cleaner React = easier to test
  • /expo-testing — Detox for React Native integration tests
  • /design-principle-enforcer — SOLID code is testable code

Recommended Test Distribution

For a typical web app with API + frontend:

Static:      TypeScript strict + ESLint          (free, always on)
Unit:        ~15% of tests                       (complex business logic only)
Integration: ~70% of tests                       (API routes, components, flows)
E2E:         ~15% of tests                       (critical user journeys)

This gives maximum confidence with minimum maintenance burden.

平台分布

Codex

35.11%
按下载量换算22

Claude

31.24%
按下载量换算20

Cursor

18.07%
按下载量换算11

Gemini CLI

8.49%
按下载量换算5

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

操作浏览器

该 Skill 可能涉及浏览器控制能力,使用时可能读取或操作网页内容,需要在受控环境中确认权限边界。

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