projection-patterns

Project event streams into read models for CQRS and analytics.

20|3|Updated Jul 26, 2025

Search Tags:#cqrs #event-sourcing #read-models #event-store #projector #stream-processing #projections

Ready to run

npx skills add https://github.com/thc1006/nephoran-intent-operator --skill projection-patterns-thc1006

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: projection-patterns
Source: https://github.com/thc1006/nephoran-intent-operator/tree/main/.claude/skills/projection-patterns
Command: npx skills add https://github.com/thc1006/nephoran-intent-operator --skill projection-patterns-thc1006

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Projections transform event streams into queryable read models, enabling CQRS-style reads, materialized views, and scalable analytics over evolving data.

Core Features & Use Cases

Projection architecture and patterns for translating events into read models
Live, catchup, persistent, and inline projection types for various consistency and performance needs
Multi-table and cross-domain read model generation to support dashboards and reporting
Templates and best practices for building robust projections with proper checkpointing and replayability
Reference strategies for implementing projections across event stores, databases, and search indexes

Quick Start

Frequently Asked Questions about projection-patterns

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I build read models from event streams for CQRS?▼

To build read models from event streams for CQRS, you project events into queryable materialized views. This requires a projection runtime with event store access to translate incoming events into structured data for dashboards and reporting.

What is the difference between live and catchup projections?▼

Live projections process events in real-time for immediate consistency, while catchup projections rebuild read models from historical event streams. Both types require checkpointing to track progress and ensure replayability across your event store.

How do I handle multi-table projections across domains?▼

Multi-table projections across domains are handled by generating cross-domain read models from your event stream. The projection runtime performs transactional updates to maintain consistency when materializing events into multiple database tables.

Do I need a database to implement event stream projections?▼

You need a projection runtime with event store access and checkpointing, but external stores like databases or search indexes are optional. Projections can materialize views directly or write to external stores for scalable analytics and reporting.

What's the best way to ensure projection replayability?▼

To ensure projection replayability, implement proper checkpointing within your projection runtime. Checkpointing tracks the event stream position, allowing catchup projections to rebuild read models deterministically from the event store without data loss.

Can I use projections for analytics and dashboards?▼

Projections are applicable for scalable analytics and dashboards by transforming event streams into materialized views. Multi-table read model generation supports reporting needs, while persistent projections maintain updated views for continuous querying.