View a markdown version of this page

Blocks reference - AWS Blocks
AuthenticationData storageReal-time and asyncAICommunicationConfigurationObservabilityHosting

Blocks reference

AWS Blocks ships with built-in Blocks covering authentication, data storage, real-time messaging, AI, observability, and more. Each Block provides a typed API, automatic resource provisioning, and a local implementation for offline development.

For more information about source code and advanced configuration, see AWS Blocks on GitHub.

Authentication

Blocks for user identity and session management.

  • AuthBasic: Username/password with JWT sessions. Use for prototypes and internal tools.

  • AuthOIDC: OIDC sign-in with Google, GitHub, Okta, or any compliant provider. Use for social login.

  • AuthCognito: Production-grade auth with MFA, social sign-in, SAML, and passkeys. Use for production applications.

For more information, see Authentication.

Data storage

Blocks for persisting and retrieving data.

  • KVStore: Simple key-value storage backed by DynamoDB. Use for caches, session stores, and feature flags.

  • DistributedTable: Structured data with schema validation, secondary indexes, and rich queries. Use for entities with multiple access patterns.

  • Database: Full PostgreSQL with Kysely query builder, migrations, transactions, and Row Level Security. Use for relational data.

  • DistributedDatabase: Serverless SQL with scale-to-zero and multi-region writes. Use for globally distributed relational data.

  • FileBucket: File storage for uploads and downloads with presigned URLs. Use for user-generated content.

For more information, see Data storage.

Real-time and async

Blocks for real-time communication and background processing.

  • Realtime: Typed WebSocket pub/sub channels. Use for live updates, chat, and notifications.

  • AsyncJob: Fire-and-forget background work via SQS and Lambda. Use for tasks that don’t need immediate results.

  • CronJob: Scheduled task execution via EventBridge. Use for periodic maintenance and batch processing.

For more information, see Real-time and async.

AI

Blocks for AI-powered features.

  • Agent: AI agent with streaming, tool calling, HITL approval, and conversation persistence. Locally, uses a canned keyword-based provider (no real model needed). On AWS, connects to Amazon Bedrock.

  • KnowledgeBase: Semantic document retrieval via Amazon Bedrock Knowledge Bases. Use for RAG and contextual search.

For more information, see AI.

Communication

Blocks for sending messages to users.

  • EmailClient: Transactional email sending via Amazon SES. Locally, captures emails for testing.

For more information, see Communication.

Configuration

Blocks for application settings and secrets.

  • AppSetting: A single configuration value or secret backed by SSM Parameter Store.

For more information, see Configuration.

Observability

Blocks for monitoring, logging, and tracing.

  • Metrics: Custom application metrics via CloudWatch Embedded Metric Format.

  • Logger: Structured JSON logging with levels and contextual metadata.

  • Tracer: Distributed tracing via AWS X-Ray.

  • Dashboard: Auto-generated observability dashboard from your metrics definitions.

For more information, see Observability.

Hosting

Frontend deployment (CDK layer only, import from @aws-blocks/blocks/cdk).

  • Hosting: Frontend deployment with SSR support. Auto-detects framework (Next.js, Nuxt, Astro, SPA).

For more information, see Hosting.