部署与云服务
infrastructure
Workflow for provisioning cloud infrastructure using IaC. Use when the user needs to create or modify cloud resources.
girijashankarj · 已有 AI 审核记录
安装与使用
运行命令前先阅读来源与依赖说明。复制命令不会自动安装或执行。
npx skills add girijashankarj/cursor-handbook --skill infrastructure技能说明
name: infrastructure description: Workflow for provisioning cloud infrastructure using IaC. Use when the user needs to create or modify cloud resources.
Skill: Set Up Cloud Infrastructure
Trigger
When the user needs to create or modify cloud resources.
Steps
Step 1: Define Requirements
- What resources are needed?
- Which environment? (dev, staging, production)
- Performance requirements?
- Budget constraints?
- Security requirements?
Step 2: Design Architecture
- Draw architecture diagram (Mermaid)
- Define networking (VPC, subnets, security groups)
- Plan compute resources (size, scaling)
- Plan storage (database, file storage, cache)
- Plan monitoring and alerting
Step 3: Write IaC
- Create Terraform/CDK/CloudFormation templates
- Use modules for reusable components
- Parameterize environment-specific values
- Follow naming conventions:
{project}-{env}-{resource} - Tag all resources
Step 4: Security
- Enable encryption at rest
- Enable encryption in transit
- Configure IAM with least privilege
- Set up VPC and security groups
- Enable audit logging
Step 5: Deploy
- Plan: Review changes before applying
- Apply to dev first
- Verify in dev
- Promote to staging
- Verify in staging
- Promote to production
Step 6: Monitor
- Set up health checks
- Configure alert rules
- Create dashboards
- Set up cost alerts
If a step fails
| Step | Failure | Recovery |
|---|---|---|
| Step 5 | Deploy to dev/staging fails | Run terraform destroy or equivalent; fix IaC; re-apply; never proceed to prod until dev/staging succeeds |
| Step 5 | Deploy to production fails | Rollback via IaC (revert and apply previous state); investigate; fix in dev; re-test full pipeline |
| Step 5 | Partial apply (some resources created) | Document what was created; run destroy/rollback for failed resources; fix and retry |
Never apply infrastructure changes to production without explicit user confirmation. Require confirmation before terraform apply or equivalent on prod.
Completion
Infrastructure provisioned, secured, monitored, and documented.