AgentSkillsCN

cloud-platforms

云基础设施与服务使用指南。当您需要部署到云端、选择云服务、配置基础设施,或解决特定于云环境的挑战时,可使用此指南。

SKILL.md
--- frontmatter
name: cloud-platforms
description: |
  Guide for working with cloud infrastructure and services. Use when
  deploying to cloud, selecting cloud services, configuring infrastructure,
  or solving cloud-specific challenges.

Cloud Platforms

Stage Guidance

Plan Stage

Focus: Select appropriate cloud services and design for availability, security, and cost efficiency.

Activities:

  • Identify service requirements
  • Select appropriate cloud services
  • Plan for high availability
  • Consider security and cost

Ready for Code when:

  • Service selection matches requirements
  • Availability approach planned
  • Security model defined
  • Cost controls considered

Code Stage

Focus: Implement cloud infrastructure with security best practices. Use infrastructure as code for reproducibility.

Activities:

  • Define infrastructure as code
  • Configure security groups and IAM
  • Set up monitoring and alerting
  • Implement deployment automation

Ready for Review when:

  • Multi-AZ deployment for availability
  • Security groups properly configured
  • IAM follows least privilege
  • Data encrypted at rest and in transit
  • Infrastructure defined as code

Review Stage

Focus: Validate security, availability, and operational readiness. Ensure cost controls are in place.

Activities:

  • Verify security configuration
  • Test availability and failover
  • Review cost projections
  • Validate monitoring coverage

Ready for Complete when:

  • Security review completed
  • Monitoring and alerting in place
  • Cost controls established
  • Operational runbooks exist

Reference

Service Categories

Compute

  • VMs/EC2: Full control, any workload
  • Containers/ECS/GKE: Portable, scalable applications
  • Serverless/Lambda: Event-driven, pay-per-use
  • Kubernetes: Container orchestration at scale

Storage

  • Object Storage (S3/GCS): Unstructured data, backups, static assets
  • Block Storage (EBS): VM disks, databases
  • File Storage (EFS): Shared file systems
  • Archive (Glacier): Long-term, infrequent access

Databases

  • Managed SQL (RDS/Cloud SQL): Relational, ACID transactions
  • NoSQL (DynamoDB/Firestore): Flexible schema, high scale
  • Cache (ElastiCache/Memorystore): Low-latency data access
  • Data Warehouse (Redshift/BigQuery): Analytics at scale

Messaging

  • Queues (SQS/Cloud Tasks): Decoupled processing
  • Pub/Sub (SNS/Cloud Pub/Sub): Event distribution
  • Streaming (Kinesis/Dataflow): Real-time data processing

Cloud-Native Design

Principles

  • Design for failure (everything fails eventually)
  • Use managed services when possible
  • Automate everything (infrastructure as code)
  • Monitor and alert on all services

High Availability

  • Deploy across multiple availability zones
  • Use load balancers for traffic distribution
  • Implement health checks and auto-healing
  • Design for graceful degradation

Security

  • Principle of least privilege for IAM
  • Encrypt data at rest and in transit
  • Use security groups and network policies
  • Rotate credentials regularly

Cost Optimization

  • Right-size instances based on actual usage
  • Use reserved instances for predictable workloads
  • Leverage spot/preemptible instances for fault-tolerant work
  • Set up billing alerts and budgets
  • Delete unused resources