Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →DevOps knowledge reference covering Git workflows, testing strategies, DevSecOps, release pipeline orchestration (release.yml, multi-arch images, cosign integration), CI/CD pipelines, database management, observability, and performance optimization. Use when working with Git, CI/CD, release pipelines, ghcr image publishing, testing, monitoring, or infrastructure automation.
.claude/skills/telagod-automating-devops/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-14 | ✗→✓ | ▲ Improved | -67% | 0% |
| case-21 | ✗→✓ | ▲ Improved | 17% | 0% |
| case-03 | ✓→✓ | ＝ Same ✓ | 25% | 0% |
| case-01 | ✓→✓ | ＝ Same ✓ | 5% | 0% |
| case-02 | ✓→✓ | ＝ Same ✓ | 2% | 0% |
> 判断先于执行：决定「是否做 / 选什么 / 如何取舍」（栈、方案、架构、权衡）前，先读领域判断内核 skills/_kernel/backend/SKILL.md——它管 judgment，本秘典管 execution；冲突时以内核判断为准。
| 意图 | 秘典 | 核心 | |------|------|------| | 版本控制 | git-workflow | Git 分支策略、PR 流程、rebase vs merge | | 测试 | testing | 单元/集成/E2E、TDD、覆盖率 | | 安全开发 | devsecops | CI/CD 安全、SAST/DAST、供应链 | | Release 编排 | release-pipelines | release.yml 多 job 骨架、metadata tag、cosign 集成、踩坑字典 | | 数据库 | database | SQL/NoSQL 选型、索引优化、迁移 | | 性能 | performance | Profiling、火焰图、基准/负载测试 | | 可观测 | observability | 日志/指标/追踪三支柱、SLO/SLI | | 成本 | cost-optimization | FinOps、右尺寸、Spot、自动伸缩 |
| 阶段 | 动作 | 工具示例 | |------|------|----------| | Commit | lint + unit test + SAST | ESLint、pytest、Semgrep | | Build | 构建 + 镜像打包 | Docker、Buildpacks | | Test | 集成测试 + E2E | Playwright、k6 | | Security | DAST + 依赖扫描 + 密钥检测 | OWASP ZAP、Trivy、gitleaks | | Deploy | 渐进发布(canary/blue-green) | ArgoCD Rollouts、Flagger | | Verify | 冒烟测试 + SLO 校验 | Prometheus、Grafana | | Rollback | 自动回滚(SLO 违约) | ArgoCD、Helm rollback |
自动化一切 | 快速反馈(<10min) | 主干开发短分支 | 不可变制品 | 环境即代码| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-03 | pass→pass | 12,625 | 12,418 | -2% | 1 | 1 | 0% | 2,014 | 2,510 | +25% | 0 | 0 | — |
case-22 | fail→fail | 21,748 | 28,250 | +30% | 1 | 1 | 0% | 3,469 | 5,577 | +61% | 0 | 0 | — |
case-01 | pass→pass | 11,490 | 11,377 | -1% | 1 | 1 | 0% | 1,884 | 1,985 | +5% | 0 | 0 | — |
case-02 | pass→pass | 11,254 | 8,356 | -26% | 1 | 1 | 0% | 1,814 | 1,852 | +2% | 0 | 0 | — |
case-04 | pass→pass | 18,156 | 17,701 | -3% | 1 | 1 | 0% | 2,410 | 3,418 | +42% | 0 | 0 | — |
case-05 | pass→pass | 12,269 | 13,437 | +10% | 1 | 1 | 0% | 2,023 | 2,522 | +25% | 0 | 0 | — |
case-06 | fail→fail | 13,408 | 13,995 | +4% | 1 | 1 | 0% | 2,051 | 2,692 | +31% | 0 | 0 | — |
case-07 | pass→pass | 13,642 | 8,542 | -37% | 1 | 1 | 0% | 2,063 | 1,872 | -9% | 0 | 0 | — |
case-08 | pass→pass | 13,010 | 11,567 | -11% | 1 | 1 | 0% | 1,762 | 2,303 | +31% | 0 | 0 | — |
case-09 | pass→pass | 14,930 | 13,681 | -8% | 1 | 1 | 0% | 2,337 | 2,595 | +11% | 0 | 0 | — |
case-10 | pass→pass | 5,671 | 5,722 | +1% | 1 | 1 | 0% | 855 | 1,370 | +60% | 0 | 0 | — |
case-11 | pass→pass | 14,229 | 2,594 | -82% | 1 | 1 | 0% | 2,243 | 976 | -56% | 0 | 0 | — |
case-12 | pass→pass | 10,375 | 3,184 | -69% | 1 | 1 | 0% | 1,514 | 1,086 | -28% | 0 | 0 | — |
case-13 | pass→pass | 13,818 | 4,861 | -65% | 1 | 1 | 0% | 2,157 | 1,154 | -46% | 0 | 0 | — |
case-14 | fail→pass | 18,056 | 3,344 | -81% | 1 | 1 | 0% | 3,293 | 1,071 | -67% | 0 | 0 | — |
case-15 | pass→pass | 16,235 | 3,884 | -76% | 1 | 1 | 0% | 2,362 | 1,157 | -51% | 0 | 0 | — |
case-16 | pass→pass | 10,657 | 2,889 | -73% | 1 | 1 | 0% | 1,670 | 1,005 | -40% | 0 | 0 | — |
case-17 | pass→pass | 11,593 | 2,642 | -77% | 1 | 1 | 0% | 1,759 | 959 | -45% | 0 | 0 | — |
case-18 | pass→pass | 18,228 | 3,090 | -83% | 1 | 1 | 0% | 2,771 | 1,030 | -63% | 0 | 0 | — |
case-19 | fail→fail | 10,274 | 10,849 | +6% | 1 | 1 | 0% | 1,425 | 1,918 | +35% | 0 | 0 | — |
case-20 | fail→fail | 18,807 | 19,915 | +6% | 1 | 1 | 0% | 2,833 | 3,486 | +23% | 0 | 0 | — |
case-21 | fail→pass | 17,626 | 16,130 | -8% | 1 | 1 | 0% | 2,544 | 2,974 | +17% | 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 +9 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.