rag retrieval patterns

7 articles · 15 co-occurring · 0 contradictions · 0 briefs

Retrieval is explicitly identified as a distinct layer of context engineering infrastructure

Related concepts

tool integration patterns 5 context window management 5 multi agent orchestration 4 context window optimization 4 state management 3 prompt engineering 3 tool use orchestration 2 system prompt architecture 2 retrieval augmented generation 2 memory persistence 2 testing strategies 1 state threading in agents 1 sampling and composability 1 query improvement as context optimization 1 prompt composition 1

Evidence chain (7 articles, showing 7)

Breaking Down AI Engineering: The 5 Layers of LLM Systems example_of

Retrieval is explicitly identified as a distinct layer of context engineering infrastructure

Model Context Protocol (MCP): Why 2026 Will Be the Year “AI Integration” Stops Being a Custom Project extends

Article identifies MCP as evolution beyond brittle RAG pipelines—moving from ad-hoc retrieval to standardized, reusable context connectors.

The Model Context Protocol (MCP) — A Complete Tutorial | by Dr. Nimrita Koul | Medium supports

Mentions RAG as context source alongside prompt/conversation history. MCP servers can implement RAG endpoints, making retrieval a composable capability.

@shao__meng: 开源github.com/google-labs-co…P8HT example_of

DESIGN.md is a structured retrieval format: tokens + prose can be indexed and retrieved selectively. The export to Tailwind/W3C DTCG is a form of context adaptation for different downstream systems.

The importance of context engineering in LLM applications | Abishek Viswanathan posted on the topic | LinkedIn supports

Example of 'what did I quote to X customer' directly describes need for intelligent search+parse+LLM pipeline—classic RAG architecture decision point.

Building Agentic RAG Systems with LangChain, Langraph, and Crewai | by Aishwarya Rawat | MI Pal | Medium extends

Shows RAG as a step within an agent workflow rather than standalone system, demonstrating how retrieval context integrates with agentic decision-making

@bcherny: 👋 Appreciate the feedback. supports

The monitoring approach ('Claude monitor feature usage and brainstorm improvements') suggests using past behavior as retrievable context to improve future performance

query this concept

$ db.articles("rag-retrieval-patterns")

$ db.cooccurrence("rag-retrieval-patterns")

$ db.contradictions("rag-retrieval-patterns")