des-persona-data-architect

Design data engineering architectures with reversible storage, ingestion, and layer boundaries.

2|Updated May 20, 2026
One-click install
npx skills add https://github.com/DKSang/DES-SKILL --skill des-persona-data-architect
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: des-persona-data-architect
Source: https://github.com/DKSang/DES-SKILL/tree/main/skills/des-persona-data-architect
Command: npx skills add https://github.com/DKSang/DES-SKILL --skill des-persona-data-architect

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It prevents data engineering plans from locking into fragile, irreversible architectural decisions too early by guiding agents to make storage, ingestion, and layer/contract choices that preserve correctness and enable replay.

Core Features & Use Cases

  • Architecture and layer design: Define bronze/silver/gold boundaries with explicit layer grains and contracts to keep lineage clear from source to serving.
  • Ingestion and transformation planning: Specify ingestion modes and transformation strategies that handle late data, deletes, and replay with idempotency and failure handling.
  • Decision governance and handoffs: Enforce architectural decision boundaries, ensure contracts align with governance/DQ/CI-CD gates, and hand off appropriately to DQ, Analytics, and Governance reviewers.

Quick Start

Use des-persona-data-architect to produce an end-to-end architecture and contract plan for a batch or streaming ingestion design that includes reversible tradeoffs and explicit layer grains.

Frequently Asked Questions about des-persona-data-architect

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

FAQPage Schema
How do I design a data architecture that supports replay and reversibility?

Design data architecture for reversibility by defining explicit grains and data contracts across bronze, silver, and gold layers. This approach ensures ingestion remains idempotent and handles failures, late data, and deletes without locking into fragile, irreversible decisions.

What is the best way to plan ingestion and transformation across bronze silver gold layers?

Plan ingestion and transformation across bronze, silver, and gold layers by specifying ingestion modes and transformation strategies that handle late data and deletes. Use idempotent ingestion with explicit failure and replay handling to maintain data correctness.

Why do data contracts matter in source-to-serving lifecycles?

Data contracts matter in source-to-serving lifecycles because they enforce explicit layer grains and keep lineage clear. Aligning contracts with governance, data quality, and CI-CD gates ensures production-ready constraints and proper handoffs to DQ and Analytics reviewers.

How do I handle late data and deletes in batch or streaming ingestion design?

Handle late data and deletes in batch or streaming ingestion design by specifying idempotent ingestion modes with robust failure and replay handling. This preserves correctness and architectural reversibility across your transformation pipeline.

When should I define partitioning and indexing choices in data architecture?

Define partitioning and indexing choices during architecture decisions for the source-to-serving lifecycle. Making these storage choices early with explicit data contracts prevents fragile pipelines and ensures transformation replay strategies remain viable.

Does des-persona-data-architect work for both batch and streaming ingestion design?

Yes, des-persona-data-architect works for both batch and streaming ingestion design. It produces an end-to-end architecture and contract plan that includes reversible tradeoffs, explicit layer grains, and idempotent ingestion handling for either mode.