Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    AWS Serverless: Lambda, DynamoDB & Events Done Right

    2

    AWS Serverless makes your agent build and review like a seasoned AWS architect.

    Free

    10 uses5.0 (1 review)
    AWS Serverless: Lambda, DynamoDB & Events Done Right

    AWS Serverless: Lambda, DynamoDB & Events Done Right

    Example session with this skill installed

    Review our serverless backend (CDK, TypeScript). We get random duplicate orders, some messages disappear, and the AWS bill doubled last month.

    • Read your context and instructions
    • Compiled the aws serverless

    Architecture found

    API Gateway REST -> createOrder Lambda -> (sync invoke) chargePayment Lambda -> (sync invoke) notify Lambda; SQS fulfilment -> worker Lambda; DynamoDB orders (provisioned). 14 functions, x86, 1,024 MB each.

    Findings (severity-ranked)

    #SeverityFindingEvidenceFix
    1criticalDuplicate orders: client retries on 504 re-create orders; no idempotencycreateOrder.ts:41, API GW 29 s timeout vs 3 chained LambdasIdempotency key header + Powertools Idempotency (DynamoDB table)
    2criticalMessages lost: SQS has no DLQ; worker throws on a bad record, so the whole batch retries until expiryqueue.ts, no deadLetterQueue, no partial batch responseDLQ (maxReceiveCount 5) + reportBatchItemFailures + DLQ alarm
    3highVisibility timeout 30 s = function timeout 30 s -> messages reprocessed while still runningqueue.ts:12180 s (6x)
    4highSync Lambda -> Lambda chain doubles duration cost and latencycreateOrder.ts:58Step Functions Express or EventBridge event
    5highCost: CloudWatch Logs = 41 % of bill (DEBUG level, never-expire retention); NAT Gateway for DynamoDB trafficCost ExplorerINFO level, 30-day retention, DynamoDB gateway endpoint
    6mediumdynamodb:* on * in 9 rolesIAMScoped grants via CDK grantReadWriteData
    7mediumScan on orders list endpointlistOrders.ts:20GSI ORG#id / created_at + Query

    Applied (PR-ready, dev stage)

    • Idempotency on createOrder, payment and notify moved to a Step Functions Express workflow
    • DLQ + partial batch failures + alarms; visibility 180 s
    • arm64 + Power Tuning: memory 1,024 -> 512 MB on 11 functions with no latency change
    • Log retention 30 days, INFO level; DynamoDB gateway endpoint
    • cdk synth clean, 38 unit tests passing, deployed to dev

    Estimated monthly cost

    $1,940 -> ~$870 (-55 %), mainly logs, NAT and Lambda duration.

    Not touched: production stack and shared IAM - deploy after your review.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Serverless that survives production (and the bill)

    The usual serverless problems: Lambdas calling Lambdas, * IAM policies, SQS without DLQs, one bad message blocking a batch, scans on DynamoDB, no alarms, surprise CloudWatch and NAT costs. AWS Serverless makes your agent build and review like a seasoned AWS architect.

    What it does

    • Uses your IaC: CDK, SAM, SST, Serverless Framework or Terraform - never adds a second one.
    • Right pattern for each job: HTTP API, API -> SQS for async work, EventBridge/SNS fan-out, Step Functions for workflows, Scheduler for cron, S3 -> SQS for files, Fargate for > 15 min tasks; service integrations instead of glue Lambdas; Mermaid diagram with sync/async and retry behaviour.
    • DynamoDB from access patterns: key and GSI design so every hot path is a Query, conditional writes, TTL, on-demand vs provisioned.
    • Reliability: idempotency, DLQs with alarms, visibility timeout >= 6x function timeout, partial batch responses, timeouts and retries with jitter, concurrency limits to protect downstreams.
    • Security: least-privilege IAM per function, Secrets Manager/SSM, auth at the edge, validation, private buckets, restricted CORS.
    • Observability: Powertools Logger, Metrics, Tracer; alarms on errors, throttles, DLQ depth, 5xx and p99; log retention.
    • Cost & cold starts: arm64, memory power tuning, batching, log costs, NAT vs VPC endpoints, bundle size, SnapStart or provisioned concurrency only where needed.
    • Safe delivery: synth/validate, tests, dev-stage deploys; never touches production or shared IAM without approval.

    Includes

    • SKILL.md workflow
    • references/patterns.md - pattern table, single-table DynamoDB example, SQS + Powertools batch handler, CDK queue/DLQ/alarm snippet
    • references/ops.md - review checklist, cost levers, cold-start tips, Powertools features

    Works with Claude Code, Codex, Cursor, Copilot, Gemini CLI and any SKILL.md-compatible agent.

    Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for APIs & Backend.

    No production deployments or resource deletions without approval. Requires user to provide existing IaC framework context to avoid tool mismatch.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before buying a skill.

    Reviews

    5.0
    1 review
    5
    1
    4
    0
    3
    0
    2
    0
    1
    0

    This skill has been used 10 times.

    Trust & safety

    Security scan

    Security scan passed 9 days ago

    • Free with an Agensi account

    Permissions

    No special permissions declared or detected

    Listed9 days ago
    Updated6 days ago

    What's inside

    Frequently Asked Questions

    This skill implements advanced production-grade patterns like DynamoDB single-table design, partial batch failure handling for SQS, and least-privilege IAM scoping that standard models often overlook.

    The skill is compatible with AWS CDK, SAM, SST, Terraform, and the Serverless Framework; it adapts its output based on the specific IaC tooling detected in your project.

    You receive the full logic package for architecting serverless systems, including patterns for high-scale event routing, cost-optimized data modeling, and reliability templates.

    Yes, the skill is designed to analyze your existing access patterns and primary keys to suggest optimized GSI strategies or TTL configurations to improve performance and lower costs.

    The skill receives regular updates to reflect the latest AWS best practices, such as new Lambda runtimes, Graviton architecture optimizations, and updated SDK versions.

    For security reasons, this skill generates the necessary IaC and handler code but will not execute deployments or stack deletions directly in your AWS account.

    New in APIs & Backend

    Free