Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →测试执行方法，包含测试框架检测、测试运行、结果解析
.claude/skills/echovic-qa-test-execution/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-16 | ✗→✓ | ▲ Improved | -47% | 0% |
| case-02 | ✓→✗ | ▼ Worse | 5% | 0% |
| case-04 | ✓→✓ | ＝ Same ✓ | 20% | 0% |
| case-05 | ✓→✓ | ＝ Same ✓ | -3% | 0% |
| case-06 | ✓→✓ | ＝ Same ✓ | 23% | 0% |
你必须真正执行测试，禁止生成 Mock 数据！
jest.config.js, "jest" in package.jsonvitest.config.js, "vitest" in package.jsonplaywright.config.jscypress.jsonpytest.ini, "pytest" in dependencies*_test.go 文件| 语言 | 单元测试 | E2E测试 | |------|----------|---------| | Node.js | npm test | npx playwright test | | Python | pytest | pytest tests/e2e | | Go | go test ./... | - |
> 完整方法论：详见 Skill(skill: "qa/e2e-playwright")
检查以下标志确认项目使用 Playwright：
playwright.config.ts 或 playwright.config.js 存在package.json 中包含 @playwright/test 依赖e2e/ 或 tests/e2e/ 目录存在bash# 安装浏览器（首次或 CI 环境） npx playwright install --with-deps # 运行全部 E2E 测试 npx playwright test # 仅 critical 标签（门禁加速） npx playwright test --grep @critical # 指定浏览器 npx playwright test --project=chromium # JSON 报告（门禁解析用） npx playwright test --reporter=json
Playwright JSON 报告关键字段：
| 字段 | 说明 | |------|------| | stats.expected | 通过的测试数 | | stats.unexpected | 失败的测试数 | | stats.flaky | 重试后通过的测试数 | | stats.skipped | 跳过的测试数 |
bash# 查看 trace（失败时自动生成） npx playwright show-trace <trace.zip路径> # 打开 HTML 报告 npx playwright show-report
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 8,631 | 4,879 | -43% | 1 | 1 | 0% | 178 | 826 | +364% | 0 | 0 | — |
case-02 | pass→fail | 5,210 | 20,025 | +284% | 1 | 1 | 0% | 794 | 836 | +5% | 0 | 0 | — |
case-03 | fail→fail | 4,357 | 4,536 | +4% | 1 | 1 | 0% | 219 | 777 | +255% | 0 | 0 | — |
case-04 | pass→pass | 5,377 | 2,626 | -51% | 1 | 1 | 0% | 801 | 963 | +20% | 0 | 0 | — |
case-05 | pass→pass | 14,483 | 10,399 | -28% | 1 | 1 | 0% | 2,544 | 2,461 | -3% | 0 | 0 | — |
case-06 | pass→pass | 5,566 | 3,023 | -46% | 1 | 1 | 0% | 873 | 1,070 | +23% | 0 | 0 | — |
case-07 | pass→pass | 2,807 | 2,609 | -7% | 1 | 1 | 0% | 298 | 984 | +230% | 0 | 0 | — |
case-08 | pass→pass | 5,711 | 2,599 | -54% | 1 | 1 | 0% | 878 | 977 | +11% | 0 | 0 | — |
case-09 | pass→pass | 4,657 | 2,731 | -41% | 1 | 1 | 0% | 746 | 1,013 | +36% | 0 | 0 | — |
case-10 | pass→pass | 8,043 | 2,677 | -67% | 1 | 1 | 0% | 1,174 | 1,021 | -13% | 0 | 0 | — |
case-11 | pass→pass | 6,675 | 3,125 | -53% | 1 | 1 | 0% | 993 | 1,079 | +9% | 0 | 0 | — |
case-12 | pass→pass | 4,254 | 1,737 | -59% | 1 | 1 | 0% | 634 | 820 | +29% | 0 | 0 | — |
case-13 | pass→pass | 8,143 | 3,513 | -57% | 1 | 1 | 0% | 1,397 | 1,168 | -16% | 0 | 0 | — |
case-14 | pass→pass | 7,395 | 2,212 | -70% | 1 | 1 | 0% | 1,186 | 919 | -23% | 0 | 0 | — |
case-15 | pass→pass | 4,840 | 2,009 | -58% | 1 | 1 | 0% | 840 | 872 | +4% | 0 | 0 | — |
case-16 | fail→pass | 10,980 | 2,373 | -78% | 1 | 1 | 0% | 1,690 | 897 | -47% | 0 | 0 | — |
case-17 | pass→pass | 2,907 | 1,869 | -36% | 1 | 1 | 0% | 463 | 899 | +94% | 0 | 0 | — |
case-18 | pass→pass | 5,344 | 1,714 | -68% | 1 | 1 | 0% | 796 | 773 | -3% | 0 | 0 | — |
case-19 | pass→pass | 5,520 | 2,720 | -51% | 1 | 1 | 0% | 941 | 993 | +6% | 0 | 0 | — |
case-20 | pass→pass | 4,362 | 3,104 | -29% | 1 | 1 | 0% | 731 | 1,051 | +44% | 0 | 0 | — |
case-21 | pass→pass | 6,031 | 4,525 | -25% | 1 | 1 | 0% | 1,029 | 1,361 | +32% | 0 | 0 | — |
case-22 | pass→pass | 2,905 | 3,149 | +8% | 1 | 1 | 0% | 485 | 1,123 | +132% | 0 | 0 | — |
case-23 | pass→pass | 5,130 | 6,657 | +30% | 1 | 1 | 0% | 928 | 1,977 | +113% | 0 | 0 | — |
case-24 | pass→pass | 15,558 | 17,029 | +9% | 1 | 1 | 0% | 2,712 | 3,822 | +41% | 0 | 0 | — |
case-25 | fail→fail | 7,522 | 9,704 | +29% | 1 | 1 | 0% | 1,541 | 2,608 | +69% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 25 cases were attempted, and 22 counted toward the lift figure. The other 3 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of 0 percentage points is the difference between those two pass rates over the 22 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.