RAG Architect skill: what it does and how to install it

What it does. Guides requirements analysis, vector-store and schema selection, document chunking, embedding and indexing, hybrid retrieval, reranking, and evaluation. It produces an architecture covering ingestion and retrieval pipelines, trade-offs, chunking rationale, query flow, and metrics with thresholds.

When to use it. For RAG, vector search, knowledge bases, semantic retrieval, and context-augmented AI applications. It applies when designing or implementing retrieval systems and covers multi-tenant filtering, deduplication, monitoring, and embedding migration.