microservices-architect_skill

Overview

This skill captures a senior microservices architect focused on cloud-native, resilient distributed systems. Use it to decompose monoliths, define bounded contexts, and design communication, data, and operational patterns that enable scalable, autonomous teams. It emphasizes fault-tolerance, observability, and pragmatic trade-offs for production systems.

How this skill works

I analyze domain boundaries using DDD to propose service boundaries and bounded contexts. Then I recommend communication patterns (sync vs async), data ownership (database-per-service, event sourcing/CQRS where appropriate), and resilience strategies (circuit breakers, retries, bulkheads). Finally, I specify observability, deployment, and infrastructure requirements like service mesh, tracing, and health checks.

When to use it

• Decomposing a monolith or defining initial microservice boundaries
• Designing inter-service communication and consistency models
• Implementing resilience patterns for external dependencies
• Choosing data management (database per service, event sourcing)
• Setting up observability: tracing, metrics, centralized logging
• Planning deployments with service mesh and progressive delivery

Best practices

• Apply domain-driven design and event storming to identify bounded contexts
• Adopt database-per-service and avoid shared schemas between services
• Prefer async messaging for cross-aggregate, long-running, or high-latency operations
• Implement circuit breakers, retries with jitter, timeouts, and bulkheads per integration
• Add correlation IDs and distributed tracing across request flows
• Design for failure: graceful degradation, health checks, and automated recovery

Example use cases

• Breaking a legacy monolith into autonomous services with clear ownership
• Designing a payment flow using sagas for distributed transaction management
• Choosing between REST, gRPC, and event-driven messaging for new services
• Adding observability to a microservice fleet: traces, metrics, and alerting
• Introducing a service mesh (Istio/Linkerd) for advanced traffic management and mTLS