skills.rest
agentsope
agentsopeCOMMUNITY
Explore all skills by this author →
GitHub

agentsop-llamaindex

Build, debug, and evaluate LlamaIndex RAG systems for grounded document answers.

287|16|Updated May 20, 2026
Search Tags:#rag#evaluation#retrieval#hybrid search#chunking#llamaindex#production hardening
One-click install
npx skills add https://github.com/agentsope/SkillAlchemy --skill agentsop-llamaindex
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: agentsop-llamaindex
Source: https://github.com/agentsope/SkillAlchemy/tree/main/skills/agentsop-llamaindex
Command: npx skills add https://github.com/agentsope/SkillAlchemy --skill agentsop-llamaindex

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This skill helps agents design and debug LlamaIndex-based RAG systems reliably, turning messy document corpora into grounded retrieval answers while preventing common ingestion, retrieval, and evaluation failure modes.

Core Features & Use Cases

  • LlamaIndex mental model distillation: Converts LlamaIndex’s Documents → Nodes → Index → Retriever → Query Engine / Response Synthesizer architecture into an actionable decision framework.
  • RAG bootstrap SOP + hardening playbook: Provides a step-by-step protocol for baseline indexing, eval-loop setup, and iterative optimization (chunking, embeddings, hybrid retrieval, reranking, routing, and production hardening).
  • Dilemma-driven decision support: Covers key architecture tradeoffs such as chunk precision vs context, hybrid BM25+dense vs dense-only, router vs agent vs decomposition, long-context vs RAG, and sentence-window vs auto-merging.

Quick Start

Tell an AI agent: "Please activate the agentsop-llamaindex skill to design a production-ready RAG pipeline in LlamaIndex for my private documents, include an eval loop, and explain which index/retrieval primitives to use for my query types."

Frequently Asked Questions about agentsop-llamaindex

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

FAQPage Schema
How do I build and evaluate a production-ready RAG pipeline with LlamaIndex?▼

To build and evaluate a LlamaIndex RAG pipeline, apply a 5-layer mental model spanning Documents to Response Synthesizers, then follow a baseline-through-eval SOP for iterative chunking, embedding, and retrieval optimization.

What is the best way to harden retrieval quality in a RAG system?▼

Harden RAG retrieval quality by applying a failure-mode checklist and iterating through an evaluation loop that tests hybrid search, reranking, and chunking strategies to ensure grounded answers from unstructured documents.

Should I use LlamaIndex or LangChain for my retrieval-augmented generation system?▼

Choose LlamaIndex over LangChain or raw vector stores when you need structured primitives for indexing and retrieval, utilizing dilemma-based decision logic to evaluate framework tradeoffs for your specific query types.

When should I use hybrid BM25+dense search versus dense-only retrieval in LlamaIndex?▼

Use hybrid BM25+dense search over dense-only retrieval when your RAG pipeline requires matching specific keywords alongside semantic context, resolving this tradeoff through the skill's dilemma-driven decision support logic.

How do I decide between using a router, an agent, or query decomposition for RAG?▼

Decide between routing, agents, or query decomposition by evaluating architecture tradeoffs like context precision versus chunk context, using the provided decision logic to match your production readiness requirements.

Why does my LlamaIndex RAG pipeline return ungrounded answers from private documents?▼

Ungrounded RAG answers often stem from ingestion or retrieval failure modes; debug them by checking the pipeline against a failure-mode checklist and verifying node chunking and synthesizer configurations.

Related Skills

Explore More Software Engineering Skills→
attentiondotnetattentiondotnet

chroma

Store, index, and search embeddings with an open-source vector database.

Community
Advanced
vasilyu1983vasilyu1983
73

ai-rag

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

Community
Advanced
neverinfamousneverinfamous
19

rag-pipelines

Design RAG pipelines with chunking, hybrid search, and cross-encoder reranking.

Community
Advanced
amanhsnamanhsn

RAG Architect

Design and optimize production-grade RAG pipelines with retrieval and evaluation strategies.

Community
Advanced