Install any skill in seconds. Free to start, no credit card required.
Get Started Free →部署流程和CI/CD配置,确保安全可靠的部署
.claude/skills/echovic-devops-deployment-process/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -8% | 0% |
| case-05 | ✓→✓ | = Same ✓ | -18% | 0% |
| case-03 | ✓→✓ | = Same ✓ | -35% | 0% |
| case-04 | ✓→✓ | = Same ✓ | -31% | 0% |
| case-06 | ✓→✓ | = Same ✓ | -28% | 0% |
| 策略 | 说明 | 适用场景 | |------|------|----------| | 蓝绿部署 | 两套环境切换 | 需要快速回滚 | | 滚动部署 | 逐步替换实例 | 零停机部署 | | 金丝雀部署 | 小流量验证 | 风险较高的更新 |
代码提交 → 自动构建 → 自动测试 → 部署到测试环境 → 部署到生产环境yamlname: CI/CD on: [push] jobs: build: runs-on: ubuntu-latest steps: - uses: actions/checkout@v2 - name: Install run: npm install - name: Test run: npm test - name: Build run: npm run build - name: Deploy run: npm run deploy
| 环境 | 用途 | 数据 | |------|------|------| | local | 本地开发 | 测试数据 | | dev | 开发测试 | 测试数据 | | staging | 预发布 | 生产数据副本 | | prod | 生产环境 | 真实数据 |
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-05 | pass→pass | 11,586 | 6,849 | -41% | 1 | 1 | 0% | 1,755 | 1,432 | -18% | 0 | 0 | — |
case-01 | fail→pass | 25,973 | 19,595 | -25% | 1 | 1 | 0% | 3,581 | 3,285 | -8% | 0 | 0 | — |
case-02 | fail→fail | 21,765 | 21,388 | -2% | 1 | 1 | 0% | 3,671 | 4,020 | +10% | 0 | 0 | — |
case-03 | pass→pass | 19,123 | 10,766 | -44% | 1 | 1 | 0% | 2,739 | 1,787 | -35% | 0 | 0 | — |
case-04 | pass→pass | 18,655 | 12,731 | -32% | 1 | 1 | 0% | 3,066 | 2,126 | -31% | 0 | 0 | — |
case-06 | pass→pass | 7,835 | 4,279 | -45% | 1 | 1 | 0% | 1,250 | 903 | -28% | 0 | 0 | — |
case-07 | fail→fail | 8,194 | 5,511 | -33% | 1 | 1 | 0% | 1,254 | 1,118 | -11% | 0 | 0 | — |
case-08 | pass→pass | 10,846 | 2,385 | -78% | 1 | 1 | 0% | 1,672 | 603 | -64% | 0 | 0 | — |
case-09 | pass→pass | 12,007 | 6,136 | -49% | 1 | 1 | 0% | 1,953 | 1,259 | -36% | 0 | 0 | — |
case-10 | pass→pass | 9,171 | 2,120 | -77% | 1 | 1 | 0% | 1,402 | 564 | -60% | 0 | 0 | — |
case-11 | pass→pass | 17,056 | 12,306 | -28% | 1 | 1 | 0% | 2,425 | 2,111 | -13% | 0 | 0 | — |
case-12 | pass→pass | 19,058 | 2,455 | -87% | 1 | 1 | 0% | 2,645 | 673 | -75% | 0 | 0 | — |
case-13 | pass→pass | 20,673 | 16,122 | -22% | 1 | 1 | 0% | 2,884 | 2,587 | -10% | 0 | 0 | — |
case-14 | pass→pass | 7,996 | 2,309 | -71% | 1 | 1 | 0% | 1,212 | 641 | -47% | 0 | 0 | — |
case-15 | pass→pass | 7,961 | 2,914 | -63% | 1 | 1 | 0% | 1,154 | 749 | -35% | 0 | 0 | — |
case-16 | pass→pass | 7,627 | 3,298 | -57% | 1 | 1 | 0% | 1,064 | 730 | -31% | 0 | 0 | — |
case-17 | pass→pass | 5,418 | 2,754 | -49% | 1 | 1 | 0% | 754 | 750 | -1% | 0 | 0 | — |
case-18 | fail→fail | 2,340 | 2,115 | -10% | 1 | 1 | 0% | 373 | 557 | +49% | 0 | 0 | — |
case-19 | fail→fail | 28,143 | 13,003 | -54% | 1 | 1 | 0% | 2,112 | 2,255 | +7% | 0 | 0 | — |
case-20 | pass→pass | 11,422 | 7,626 | -33% | 1 | 1 | 0% | 1,998 | 1,559 | -22% | 0 | 0 | — |
case-21 | pass→pass | 9,681 | 7,069 | -27% | 1 | 1 | 0% | 1,755 | 1,694 | -3% | 0 | 0 | — |
case-22 | pass→pass | 7,462 | 4,800 | -36% | 1 | 1 | 0% | 1,353 | 1,092 | -19% | 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. 22 cases were attempted. The headline lift of +5 percentage points is the difference between those two pass rates over the 22 comparable cases.
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.