skills.rest
aisa-group
aisa-groupOFFICIAL
Explore all skills by this author →
WebsiteGitHub

rag-implementation

Build retrieval-augmented generation systems with embeddings, vector stores, and reranking.

90|4|Updated Feb 22, 2026
Search Tags:#embeddings#rag#retrieval#reranking#semantic-search#vector-database#document-qa
One-click install
npx skills add https://github.com/aisa-group/skill-inject --skill rag-implementation-aisa-group
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: rag-implementation
Source: https://github.com/aisa-group/skill-inject/tree/main/data/skills/rag-implementation
Command: npx skills add https://github.com/aisa-group/skill-inject --skill rag-implementation-aisa-group

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

Retrieval-Augmented Generation (RAG) reduces hallucinations and enables LLMs to answer questions accurately by grounding responses in external documents and knowledge stores.

Core Features & Use Cases

  • Embeddings & Vector Stores: Guidance for selecting and configuring embeddings and vector databases (Chroma, Pinecone, Weaviate, Milvus, FAISS).
  • Retrieval & Reranking Pipelines: Patterns for dense/sparse/hybrid retrieval, multi-query strategies, contextual compression, and cross-encoder reranking to improve result relevance.
  • Use Case: Build an internal documentation Q&A chatbot that returns cited source passages and confidence scores when answering employee questions.

Quick Start

Use the rag-implementation skill to build a retrieval pipeline that indexes your docs directory, creates embeddings, and answers a natural language question with cited source passages.

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 retrieval-augmented generation system for an enterprise knowledge base?▼

To build a retrieval-augmented generation system, you create a pipeline that indexes documents, generates embeddings, stores them in a vector database, and retrieves relevant passages to ground LLM responses and reduce hallucinations.

What's the best way to improve retrieval relevance with embeddings and vector stores?▼

Improving retrieval relevance involves applying cross-encoder reranking and hybrid retrieval strategies on top of your initial embeddings to filter and prioritize the most relevant document chunks from your vector store.

How does document chunking affect semantic search over proprietary corpora?▼

Document chunking splits large texts into smaller passages before embedding generation, directly impacting semantic search accuracy by ensuring the vector database retrieves highly specific context for the LLM.

Can I use Pinecone or Weaviate for a document Q&A chatbot?▼

Yes, you can configure Pinecone or Weaviate as your vector database for a document Q&A chatbot. The system supports integrating these stores to manage embeddings and return cited source passages.

Why does my LLM hallucinate answers instead of using external knowledge?▼

LLMs hallucinate when lacking grounded context; retrieval-augmented generation fixes this by fetching external documents via semantic search and passing the retrieved passages as context to anchor the model's answers.

Do I need a cross-encoder reranker for my RAG pipeline?▼

A cross-encoder reranker is needed when initial dense or sparse retrieval returns too many irrelevant chunks; it significantly improves result relevance by re-scoring retrieved passages before sending them to the LLM.

Related Skills

Explore More Software Engineering Skills→
dotlab-hqdotlab-hq

langchain-rag

Build RAG pipelines that ingest, chunk, embed, and retrieve documents for LLMs.

Official
Intermediate
milasaurusmilasaurus

ml-rag

Build and query a retrieval-augmented generation pipeline from documents.

Community
Advanced
alexsandrocruzalexsandrocruz

prompt-engineering-patterns

Design production prompt templates with few-shot and chain-of-thought patterns.

Community
Advanced
giuseppe-trisciuogliogiuseppe-trisciuoglio
322

rag-implementation

Build retrieval-augmented generation pipelines with vector databases and embedding models.

Community
Advanced