skills.rest
svssdeva
svssdevaCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

rag-implementation

Build Retrieval-Augmented Generation systems with semantic search and vector storage.

1|1|Updated Apr 20, 2026
Search Tags:#semantic search#embeddings#rag#vector database#reranking#retrieval pipeline#llm grounded answers
One-click install
npx skills add https://github.com/svssdeva/agentic-skills --skill rag-implementation-svssdeva
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: rag-implementation
Source: https://github.com/svssdeva/agentic-skills/tree/main/ai/rag-implementation
Command: npx skills add https://github.com/svssdeva/agentic-skills --skill rag-implementation-svssdeva

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps you build Retrieval-Augmented Generation systems that answer user questions using external, trusted documents instead of relying on the model’s memory and reducing hallucinations.

Core Features & Use Cases

  • Vector database + embeddings setup: Choose an appropriate vector store (e.g., Pinecone, Weaviate, Milvus, Chroma, Qdrant, pgvector) and embedding model to support semantic search.
  • Retrieval pipeline design: Implement dense, sparse, and hybrid retrieval strategies, including multi-query and HyDE-style query expansion.
  • Answer quality improvements: Add reranking and contextual compression to improve relevance, diversity, and groundedness.
  • Chunking and indexing strategies: Use recursive, token-based, semantic, or header-aware splitting to optimize what gets retrieved.
  • Production-ready patterns: Include evaluation metrics (precision/recall, faithfulness, answer relevance) and citation-oriented prompting for verifiable outputs.

Quick Start

Use the rag-implementation skill to design a RAG flow that ingests your documents, builds embeddings and a vector index, retrieves the top relevant passages for a question, reranks results, and generates a grounded answer with citations.

Frequently Asked Questions about rag-implementation

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

FAQPage Schema
How do I build a RAG pipeline that grounds LLM answers in external documents?▼

To build a RAG pipeline, ingest documents, generate embeddings, and store them in a vector database. You then retrieve top relevant passages using semantic search, apply reranking, and generate grounded answers with citations to reduce hallucinations.

What chunking strategies work best for semantic search and vector database retrieval?▼

Effective chunking strategies for semantic search include recursive, token-based, semantic, and header-aware splitting. These indexing methods optimize what gets retrieved from your vector database by preserving contextual boundaries within the source documents.

How does reranking improve retrieval quality in a retrieval-augmented generation system?▼

Reranking improves retrieval quality by reordering initially retrieved passages to maximize relevance and diversity. Combined with contextual compression, it filters out noise before passing documents to the LLM, resulting in more accurate, grounded answers.

Can I use multi-query and HyDE patterns for hybrid retrieval in document Q&A?▼

Yes, you can implement multi-query and HyDE-style query expansion to enhance hybrid retrieval. These patterns generate multiple query variations, improving the recall of relevant passages for document Q&A and semantic retrieval tasks.

What evaluation metrics should I use to measure RAG faithfulness and retrieval quality?▼

To measure RAG faithfulness and retrieval quality, use evaluation metrics such as precision, recall, answer relevance, and faithfulness. These metrics verify that generated outputs are accurate and properly grounded in retrieved documents.

Which vector stores support embeddings-backed retrieval for knowledge-grounded chatbots?▼

Vector stores like Pinecone, Weaviate, Milvus, Chroma, Qdrant, and pgvector support embeddings-backed retrieval. Choosing the appropriate vector store and embedding model establishes the semantic search foundation for knowledge-grounded chatbots.

Related Skills

Explore More Software Engineering Skills→
patricio0312revpatricio0312rev
5

vector-db-setup

Configure Pinecone, Chroma, pgvector, and Qdrant vector databases for semantic search.

Community
Advanced
lagarcesslagarcess
1

rag-architect

Design and optimize retrieval-augmented generation pipelines for document indexing and vector search.

Community
Advanced
bobmatnycbobmatnyc
3

langchain-framework

Build and orchestrate LLM-powered applications with LangChain.

Community
Intermediate
vasilyu1983vasilyu1983
73

ai-rag

Design and implement RAG retrieval systems with hybrid search and reranking.

Community
Advanced