skills.rest
JayRHa
JayRHaCOMMUNITY
Explore all skills by this author →
Website𝕏@jannik_reinhardGitHub

rag-pipeline-designer

Design end-to-end RAG pipelines with chunking, embeddings, retrieval, and evaluation.

3|2|Updated Jun 9, 2026
Search Tags:#design#embeddings#rag#evaluation#retrieval#vector-store#rag-pipeline
One-click install
npx skills add https://github.com/JayRHa/AgentSkills --skill rag-pipeline-designer
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: rag-pipeline-designer
Source: https://github.com/JayRHa/AgentSkills/tree/main/rag-pipeline-designer
Command: npx skills add https://github.com/JayRHa/AgentSkills --skill rag-pipeline-designer

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

Designs end-to-end retrieval-augmented generation (RAG) systems by making principled choices for chunking strategy, embeddings, vector indexing, retrieval, reranking, prompt assembly, and offline evaluation. Use this skill when the user wants to build or improve a RAG pipeline, asks about chunking strategy, embedding model selection, hybrid search, reranking, "my retrieval is bad / hallucinating / missing context", chunk size tuning, vector database choice, or how to measure retrieval quality (recall@k, MRR, nDCG, faithfulness).

Core Features & Use Cases

This skill turns the design of a RAG pipeline into a concrete, defensible architecture by orchestrating ingestion, chunking, embeddings, indexing, retrieval, reranking, context assembly, and evaluation. It includes guidance on selecting components, evaluating retrieval quality, and maintaining a design-doc workflow with templates and examples. Typical use cases include building document search over large corpora, multi-hop QA pipelines, and evaluation-ready RAG prototypes.

Quick Start

Characterize the corpus, constraints, and evaluation goals; draft a design doc; and begin with a small prototype using the provided templates and scripts.

Frequently Asked Questions about rag-pipeline-designer

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

FAQPage Schema
How do I design a RAG pipeline for document search and multi-hop QA?▼

Designing a RAG pipeline requires selecting chunking strategies, embedding models, vector stores, and retrieval workflows. This skill orchestrates ingestion, indexing, reranking, context assembly, and offline evaluation into a concrete, defensible architecture.

Why does my RAG pipeline hallucinate or miss context during retrieval?▼

RAG pipelines hallucinate or miss context when chunking, embedding, or retrieval workflows are poorly configured. This skill addresses these failures by specifying reranking, context assembly, and evaluation metrics like recall@k, MRR, and faithfulness.

What is the best way to choose an embedding model and vector database for RAG?▼

Choosing embeddings and vector databases for RAG depends on corpus characteristics and performance constraints. This skill provides principled selection guidance for indexing, hybrid search, and vector stores to ensure scalable, auditable architecture.

How do I measure retrieval quality and evaluate a RAG pipeline?▼

Measure retrieval quality in a RAG pipeline using offline evaluation metrics like recall@k, MRR, nDCG, and faithfulness. This skill specifies evaluation requirements and provides templates to build evaluation-ready prototypes.

Can I tune chunk size and apply hybrid search for large corpora in RAG?▼

Tuning chunk size and applying hybrid search for large corpora are core components of a robust RAG pipeline. This skill guides chunking strategy selection and retrieval workflow design to meet performance constraints and support multi-hop queries.

What are the limitations of not using a design doc workflow for RAG architecture?▼

Without a design doc workflow, RAG architecture choices for chunking, embeddings, and retrieval lack auditability and defensibility. This skill enforces a structured design process with templates and scripts to maintain scalable, principled pipelines.

Related Skills

Explore More Software Engineering Skills→
SentioLabsSentioLabs
1

arc-brainstorm

Capture approved design approaches in docs/plans and register them with arc plan.

Official
Advanced
hamelsmuhamelsmu
1.6k

evaluate-rag

Evaluate RAG pipelines by measuring retrieval metrics and generation quality.

Community
Advanced
patornpatorn

langchain-rag

Build end-to-end RAG pipelines with LangChain document loaders, embeddings, and vector stores.

Community
Advanced
OpenRouterTeamOpenRouterTeam
209

create-agent-tui

Scaffold a TypeScript terminal agent TUI project with OpenRouter model calling and tool execution.

Official
Advanced