How do I install the RAG Design skill?

Run npx skillmds@latest add tomevault-io/rag-design in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

What does the RAG Design skill do?

Design a RAG architecture for a use case Use when this capability is needed. It is listed under AI & ML on SkillMD.

Is RAG Design safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with RAG Design?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is RAG Design free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published RAG Design?

tomevault-io (@tomevault-io) published this skill. Their other Agent Skills are listed on their SkillMD profile.

RAG Design by tomevault-io — Agent Skill

Name: RAG Design
Author: tomevault-io

File contents

Design RAG Architecture

Design a Retrieval-Augmented Generation system for a given use case.

Arguments

$ARGUMENTS - The RAG use case to design for (e.g., "customer support chatbot", "documentation Q&A", "legal document search", "code assistant")

Workflow

Clarify requirements by understanding:
- What type of questions will be asked?
- What is the document corpus size and type?
- What is the required accuracy/faithfulness?
- What is the latency budget?
- Are there multi-turn conversation requirements?
Load relevant skills based on the use case:
- RAG patterns → rag-architecture
- Vector store selection → vector-databases
- LLM serving → llm-serving-patterns
- Inference optimization → ml-inference-optimization
Spawn the rag-architect agent for comprehensive design:
- Use Task tool with subagent_type="rag-architect"
- Provide full use case context and requirements
- Request end-to-end RAG architecture
Design the ingestion pipeline:
- Document extraction (PDF, HTML, code)
- Chunking strategy selection
- Embedding model selection
- Vector database configuration
- Metadata extraction and indexing
Design the retrieval pipeline:
- Query processing (expansion, HyDE)
- Retrieval strategy (dense, sparse, hybrid)
- Reranking approach
- Context assembly
- Prompt engineering
Address quality and scale:
- Retrieval accuracy (recall@k, MRR)
- Answer faithfulness (grounding)
- Latency budget allocation
- Cost optimization
- Scaling strategy

Example Usage

/sd:rag-design customer support chatbot with 10K FAQ documents
/sd:rag-design internal documentation Q&A for engineering team
/sd:rag-design legal document search for contract review
/sd:rag-design code assistant for enterprise codebase
/sd:rag-design research paper Q&A with 100K papers
/sd:rag-design product catalog search with structured data
/sd:rag-design multi-lingual knowledge base

Use Case Categories

Category	Key Considerations
Customer Support	FAQ coverage, escalation, tone consistency
Documentation	Technical accuracy, code examples, versioning
Legal/Compliance	Citation accuracy, audit trails, access control
Code Assistance	AST-aware chunking, context relevance, IDE integration
Research/Academic	Multi-document reasoning, citation, long-form answers
E-commerce	Product attributes, inventory awareness, personalization

RAG Pattern Selection Guide

Complexity	Pattern	When to Use
Low	Basic RAG	Simple Q&A, small corpus
Medium	RAG + Reranking	Higher accuracy needed
Medium	Hybrid Search	Mixed keyword + semantic queries
High	Query-Transformed	Vague or complex queries
High	Agentic RAG	Multi-hop reasoning, tool use

Output

A comprehensive RAG system architecture including:

Ingestion pipeline (documents → vectors)
Retrieval pipeline (query → context)
Technology stack (embedding model, vector DB, LLM)
Quality targets (recall, faithfulness, latency)
Trade-offs and alternatives
Cost estimate (per-query and monthly)

Converted and distributed by TomeVault — claim your Tome and manage your conversions.

---name: rag-designdescription: Design a RAG architecture for a use case Use when this capability is needed.---# Design RAG ArchitectureDesign a Retrieval-Augmented Generation system for a given use case.## Arguments`$ARGUMENTS` - The RAG use case to design for (e.g., "customer support chatbot", "documentation Q&A", "legal document search", "code assistant")## Workflow1. **Clarify requirements** by understanding: - What type of questions will be asked? - What is the document corpus size and type? - What is the required accuracy/faithfulness? - What is the latency budget? - Are there multi-turn conversation requirements?2. **Load relevant skills** based on the use case: - RAG patterns → `rag-architecture` - Vector store selection → `vector-databases` - LLM serving → `llm-serving-patterns` - Inference optimization → `ml-inference-optimization`3. **Spawn the rag-architect agent** for comprehensive design: - Use Task tool with subagent_type="rag-architect" - Provide full use case context and requirements - Request end-to-end RAG architecture4. **Design the ingestion pipeline**: - Document extraction (PDF, HTML, code) - Chunking strategy selection - Embedding model selection - Vector database configuration - Metadata extraction and indexing5. **Design the retrieval pipeline**: - Query processing (expansion, HyDE) - Retrieval strategy (dense, sparse, hybrid) - Reranking approach - Context assembly - Prompt engineering6. **Address quality and scale**: - Retrieval accuracy (recall@k, MRR) - Answer faithfulness (grounding) - Latency budget allocation - Cost optimization - Scaling strategy## Example Usage```bash/sd:rag-design customer support chatbot with 10K FAQ documents/sd:rag-design internal documentation Q&A for engineering team/sd:rag-design legal document search for contract review/sd:rag-design code assistant for enterprise codebase/sd:rag-design research paper Q&A with 100K papers/sd:rag-design product catalog search with structured data/sd:rag-design multi-lingual knowledge base```## Use Case Categories| Category | Key Considerations || -------- | ------------------ || Customer Support | FAQ coverage, escalation, tone consistency || Documentation | Technical accuracy, code examples, versioning || Legal/Compliance | Citation accuracy, audit trails, access control || Code Assistance | AST-aware chunking, context relevance, IDE integration || Research/Academic | Multi-document reasoning, citation, long-form answers || E-commerce | Product attributes, inventory awareness, personalization |## RAG Pattern Selection Guide| Complexity | Pattern | When to Use || ---------- | ------- | ----------- || Low | Basic RAG | Simple Q&A, small corpus || Medium | RAG + Reranking | Higher accuracy needed || Medium | Hybrid Search | Mixed keyword + semantic queries || High | Query-Transformed | Vague or complex queries || High | Agentic RAG | Multi-hop reasoning, tool use |## OutputA comprehensive RAG system architecture including:- Ingestion pipeline (documents → vectors)- Retrieval pipeline (query → context)- Technology stack (embedding model, vector DB, LLM)- Quality targets (recall, faithfulness, latency)- Trade-offs and alternatives- Cost estimate (per-query and monthly)---> Converted and distributed by [TomeVault](https://tomevault.io/claim/melodic-software) — claim your Tome and manage your conversions.

Frequently asked questions

RAG Design