Agent-Skills.md

Agent Skills: AWS Serverless & Event-Driven Architecture

AWS serverless and event-driven architecture expert based on Well-Architected Framework. Use when building serverless APIs, Lambda functions, REST APIs, microservices, or async workflows. Covers Lambda with TypeScript/Python, API Gateway (REST/HTTP), DynamoDB, Step Functions, EventBridge, SQS, SNS, and serverless patterns. Essential when user mentions serverless, Lambda, API Gateway, event-driven, async processing, queues, pub/sub, or wants to build scalable serverless applications with AWS best practices.

UncategorizedID: leegonzales/aiskills/aws-serverless-eda

Repository

leegonzales/aiskills
leegonzalesLicense: MIT
255

Install this agent skill to your local

pnpm dlx add-skill https://github.com/leegonzales/AISkills/tree/HEAD/AWSSkills/aws-serverless-eda

Skill Files

Browse the full folder contents for aws-serverless-eda.

Download Skill

Loading file tree…

AWSSkills/aws-serverless-eda/SKILL.md

Skill Metadata

Name
aws-serverless-eda
Description
AWS serverless and event-driven architecture expert based on Well-Architected Framework. Use when building serverless APIs, Lambda functions, REST APIs, microservices, or async workflows. Covers Lambda with TypeScript/Python, API Gateway (REST/HTTP), DynamoDB, Step Functions, EventBridge, SQS, SNS, and serverless patterns. Essential when user mentions serverless, Lambda, API Gateway, event-driven, async processing, queues, pub/sub, or wants to build scalable serverless applications with AWS best practices.

AWS Serverless & Event-Driven Architecture

Guidance for building serverless applications and event-driven architectures on AWS based on Well-Architected Framework principles.

When to Use This Skill

Use when building serverless apps with Lambda, designing event-driven architectures, implementing microservices, creating async processing workflows, orchestrating multi-service transactions, building real-time data pipelines, implementing saga patterns for distributed transactions, or designing for scale and resilience.

Integrated MCP Servers

This skill includes 5 MCP servers. Use them as the primary tools for the tasks below.

  • AWS Documentation MCP — ALWAYS verify AWS service info before implementation: latest features/best practices, regional availability, service limits/quotas, API specs and parameters.
  • AWS Serverless MCP — full serverless lifecycle with SAM CLI: initialize, generate SAM templates, deploy, test Lambda locally before deploy.
  • AWS Lambda Tool MCP — execute Lambda functions as tools: invoke directly, test integrations, run automation, access private resources.
  • AWS Step Functions MCP — orchestration: create/manage state machines, run distributed transactions, implement saga patterns, coordinate microservices.
  • Amazon SNS/SQS MCP — event-driven messaging: publish to SNS topics, send/receive SQS messages, implement pub/sub, debug message routing.

Well-Architected Serverless Design Principles

The 7 design principles that drive every decision in this skill (apply them; reference files carry the implementation patterns):

  1. Speedy, Simple, Singular — functions are concise and single-purpose. Minimize cold starts, optimize memory, reuse connections, use provisioned concurrency only when needed.
  2. Think Concurrent Requests, Not Total — Lambda scales horizontally; design for concurrent-execution limits, downstream throttling, shared-resource contention, and connection-pool sizing (e.g. DynamoDB PAY_PER_REQUEST or provisioned + auto-scaling).
  3. Share Nothing — runtime environments are short-lived; never rely on local FS (/tmp is ephemeral). Persist state in DynamoDB (data), Step Functions (workflow), ElastiCache (session), S3 (files).
  4. Assume No Hardware Affinity — be hardware-agnostic and portable; configure via environment variables, avoid hardware-specific optimizations, test across environments.
  5. Orchestrate with State Machines, Not Function Chaining — use Step Functions for sequencing, error handling/retries, execution history, and parallel/sequential flows instead of Lambda-invokes-Lambda.
  6. Use Events to Trigger Transactions — prefer event-driven (S3 notifications, EventBridge rules) over synchronous request/response for loose coupling, async processing, fault tolerance, independent scaling.
  7. Design for Failures and Duplicates — make operations idempotent (dedupe-check before processing); use retry with exponential backoff and DLQs.

Core Pattern Catalog

Pattern implementations (CDK/TypeScript) live in reference files. Reach for:

  • Event-driven: event router (EventBridge), queue-based processing (SQS), pub/sub fan-out (SNS+SQS), saga with Step Functions, event sourcing → references/eda-patterns.md
  • Serverless: API-driven microservices, stream processing (Kinesis), async task processing, scheduled jobs (EventBridge cron/rate), webhook processing → references/serverless-patterns.md

Key Operational Practices

  • Partial batch failure — return batchItemFailures from SQS/stream handlers so only failed records retry; set reportBatchItemFailures: true on the event source mapping.
  • Dead Letter Queues — always configure a DLQ (maxReceiveCount) and alarm on DLQ depth.
  • Observability — enable X-Ray (tracing: ACTIVE) and Lambda Powertools env vars (POWERTOOLS_SERVICE_NAME, POWERTOOLS_METRICS_NAMESPACE, LOG_LEVEL).

Full implementations and deeper guidance are in the reference files below.

Reference Files

  • references/serverless-patterns.md — core serverless architectures, API patterns, data processing, Step Functions orchestration, anti-patterns, plus fan-out/scheduled/webhook trigger patterns.
  • references/eda-patterns.md — event routing/processing, event sourcing, saga patterns, idempotency, ordering/deduplication, error handling.
  • references/security-best-practices.md — shared responsibility model, IAM least privilege, data protection/encryption, VPC network security.
  • references/observability-best-practices.md — metrics/logs/traces, structured logging with Powertools, X-Ray tracing, CloudWatch alarms/dashboards.
  • references/performance-optimization.md — cold-start optimization, memory/CPU tuning, package-size reduction, provisioned concurrency.
  • references/deployment-best-practices.md — CI/CD pipelines, testing (unit/integration/load), canary/blue-green strategies, rollback safety.

External Resources:

  • AWS Well-Architected Serverless Lens: https://docs.aws.amazon.com/wellarchitected/latest/serverless-applications-lens/
  • ServerlessLand.com — pre-built serverless patterns
  • AWS Serverless Workshops: https://serverlessland.com/learn?type=Workshops