systems-architect_skill

Overview

This skill designs resilient, secure, and cost-effective infrastructure, networks, and cloud systems using proven architecture patterns. It focuses on integration, reliability, observability, and disaster recovery while balancing cost and operational complexity. The guidance translates business requirements into actionable infrastructure choices and runbooks.

How this skill works

The skill inspects system requirements (SLOs, RTO/RPO, traffic patterns, and compliance) and maps them to architecture patterns: multi-AZ/region deployments, network segmentation, and service integration choices. It recommends concrete infrastructure-as-code practices, failure modes, and operational controls (monitoring, chaos testing, DR drills). It also produces migration patterns, capacity plans, and security controls tailored to the environment.

When to use it

• Designing a new cloud-native system or migrating a legacy application to the cloud
• Defining availability, recovery, and security requirements for business-critical services
• Scaling infrastructure to support predictable growth or sudden traffic bursts
• Planning cross-account network topology and service discovery for multi-team environments
• Preparing for compliance audits or post-incident hardening and remediation

Best practices

• Design for failure at every layer; automate recovery with immutable infrastructure and IaC
• Define SLOs and error budgets before technical decisions; align capacity and cost to those targets
• Use managed services for undifferentiated work and reserve custom engineering for core differentiators
• Segment networks, apply least privilege, centralize secrets, and log everything for forensics
• Adopt metrics, logs, and distributed tracing together; alert on user-facing symptoms, not low-level causes

Example use cases

• Architecting a multi-region web service with cross-region replication and hot/warm standby for critical data
• Designing a microservices platform with service mesh, rate limiting, and dead-letter queues for integration resilience
• Creating an IaC-based migration plan using blue-green or strangler patterns and tested rollback procedures
• Building capacity plans with load testing and cost models to justify reserved vs. spot capacity
• Implementing disaster recovery drills, automated restores, and documented runbooks for 24/7 operations