🎯

microservices

🎯Skill
g1joshi/agent-skills2 Installs9Last Updated Feb 10, 2026

Installation

$npx skills add g1joshi/agent-skills --skill microservices

Run it in a terminal; it asks which AI coding tools to install it for (Claude Code, Codex, Cursor and more).

Uses skills.sh Β· Node.js 22.20+

After installing

Ask for related work and the agent uses it on its own, or call it yourself.

Run it (Claude Code)/microservices
Install by hand

Copy the whole skill folder from GitHub to the place below. Copying only SKILL.md can leave out files the skill uses.

Open the skill folder β†’
  • This project .claude/skills/microservices/
  • All projects ~/.claude/skills/microservices/

These paths are for Claude Code. For other tools, the command above puts it in the right place.

Expert Microservices architecture assistance covering service decomposition, service discovery, distributed tracing, circuit breakers, and database-per-service isolation. Use when refactoring monoliths, designing distributed microservice systems, or orchestrating service communications.

Microservices

Microservices architecture structures an application as a collection of loosely coupled, independently deployable services. Ideally, each service corresponds to a Bounded Context (DDD).

When to Use

  • Independent Scaling & Deployment: Different business capabilities (e.g. video processing vs user authentication) require independent scaling.
  • Autonomous Multi-Team Velocity: Large engineering organizations where decoupled cross-functional teams need to deploy without stepping on each other.
  • Polyglot Technology Requirements: Leveraging specific runtimes for specific workloads (Python for ML, Go/Rust for networking, Node for APIs).
  • Fault Domain Isolation: Preventing an out-of-memory crash in a non-critical feature from bringing down the core payment pipeline.

Quick Start

# docker-compose.yml (Simulated Microservices)
services:
  order-service:
    build: ./services/order
    ports: ["3001:3000"]
    environment:
      - DB_HOST=order-db

  inventory-service:
    build: ./services/inventory
    ports: ["3002:3000"]

  api-gateway:
    image: nginx
    ports: ["80:80"]
    depends_on:
      - order-service
      - inventory-service

Core Concepts

Database-Per-Service Isolation

Each microservice owns its private database. No service can directly query another service's database tables:

[ Orders Service ] ──→ ( Orders DB )
        β”‚ (Async Events / gRPC)
        β–Ό
[ Customers Service ] ──→ ( Customers DB )

Distributed Tracing (W3C Trace Context)

Correlates requests across multiple network hops using standardized trace IDs:

// Injected into outgoing HTTP/gRPC calls
headers: {
  "traceparent": "00-4bf92f3577b34da6a3ce929d0e0e4736-00f067aa0ba902b7-01"
}

Circuit Breakers & Graceful Degradation

Fails fast when downstream services become unresponsive, preventing cascading thread exhaustion:

import CircuitBreaker from "opossum";

const breaker = new CircuitBreaker(callRecommendationService, {
  timeout: 1000,
  errorThresholdPercentage: 50,
  resetTimeout: 10000,
});

breaker.fallback(() => [/* Fallback cached generic recommendations */]);
const recommendations = await breaker.fire(userId);

Common Patterns

Circuit Breaker Pattern

Problem: Cascading failures when a downstream dependency experiences degraded response times.
Solution: Wrap external calls in a circuit breaker to fail fast and shed load.

import CircuitBreaker from "opossum";

async function fetchPaymentGateway(payload: PaymentPayload) {
  return await http.post("https://payment.example.com/charge", payload);
}

const breaker = new CircuitBreaker(fetchPaymentGateway, {
  timeout: 3000, // 3s timeout
  errorThresholdPercentage: 50, // Open breaker if 50% calls fail
  resetTimeout: 10000, // Wait 10s before attempting half-open state
});

breaker.fallback(() => ({ status: "QUEUED_FOR_RETRY", cached: true }));
const result = await breaker.fire(paymentData);

Best Practices

Do:

  • Enforce Database-per-Service: Never share a database instance between microservices; communicate via APIs or events.
  • Implement Comprehensive Observability: Instrument all services with OpenTelemetry traces, Prometheus metrics, and structured JSON logs.
  • Use Contract Testing (Pact): Validate API compatibility between consumer and provider services in CI/CD without running end-to-end clusters.
  • Adopt Service Meshes for Mesh Security: Leverage Envoy / Istio for automated mTLS encryption and traffic steering.

Don't:

  • Start with microservices on day one: Start with a well-structured Modular Monolith until organizational scale warrants extraction.
  • Build distributed monoliths: Avoid deep synchronous HTTP chains (Service A calls B, which calls C, which calls D).
  • Ignore distributed transactions: Use Saga patterns or eventual consistency rather than two-phase commits.

Troubleshooting

ErrorCauseSolution
Cascading FailureOne service down takes down others.Use Circuit Breakers and Timeouts.
Data inconsistencyAsync updates failed.Implement Sagas and Idempotent consumers.
LatencyToo many hops (service -> service -> service).Use Caching, Aggregation at Gateway, or Event-Driven data replication.

References

More from this repository10