
This is a reference guide that Claude can pull from when you're working with cloud infrastructure, covering the full stack from Kubernetes manifests to GitOps automation to infrastructure as code. It includes checklists for production readiness like liveness probes, HPA, PDB setup for K8s, automated sync and secret encryption for GitOps, and state management best practices for Terraform and friends. The content is partly in Chinese but covers standard tooling like ArgoCD, Flux, Helm, Kustomize, and the major IaC frameworks. Useful when you need Claude to remember deployment patterns or catch configuration gaps, though you'll want to verify the checklist items match your team's standards.
$ npx -y skills add telagod/code-abyss --skill provisioning-infrastructure --agent claude-codeInstalls into .claude/skills of the current project.
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。
GitOps控制平面(ArgoCD/Flux) → Kubernetes(Helm/Kustomize) → IaC(Terraform/Pulumi/CDK)
livenessProbe+readinessProbe | requests+limits | HPA | PDB | ResourceQuota | 镜像 Digest | Pod 反亲和 | External Secrets
automated prune+selfHeal | ApplicationSet 多环境 | 密钥加密(Sealed Secrets/ESO) | 镜像自动更新
模块化复用 | 环境隔离不同 State | 远程状态+锁定 | Provider 版本锁 | Secrets Manager | 统一标签 | Plan 审查再 Apply
Helm/Kustomize/Operator、GitOps 工具对比、IaC 工具详情详见 references/details.md
Sponsors