rag-implementation

Build Retrieval-Augmented Generation (RAG) systems for LLM applications with vector databases and semantic search. Use when implementing knowledge-grounded AI, building document Q&A systems, or integrating LLMs with external knowledge bases.

2.12x

Quality

66%

Does it follow best practices?

Impact

70%

2.12x

Average score across 3 eval scenarios

Securityby

Passed

No known issues

Optimize this skill with Tessl

npx tessl skill review --optimize ./tests/ext_conformance/artifacts/agents-wshobson/llm-application-dev/skills/rag-implementation/SKILL.md

Evaluation results

90%

49%

Internal Documentation Q&A Assistant

LangGraph RAG pipeline with Anthropic stack and citation prompting

Criteria

Without context

With context

LangGraph StateGraph

100%

Claude model version

Voyage embedding model

100%

RAGState schema

100%

Retriever k=4

62%

100%

Local vector store

87%

100%

RecursiveCharacterTextSplitter separators

50%

100%

Chunk size range

100%

Chunk overlap ratio

100%

Citation format in prompt

37%

100%

Fallback phrase

100%

Graph edge order

100%

38%

16%

Technical Support Knowledge Base Search

Hybrid search with cross-encoder reranking and metadata filtering

Criteria

Without context

With context

EnsembleRetriever usage

70%

Hybrid search weights

Component k values

80%

Cross-encoder model

100%

Reranking candidate pool

Top-k after reranking

Metadata fields

100%

70%

Voyage embeddings

Claude model

50%

LangGraph or async pattern

84%

48%

Scientific Literature Q&A with Quality Evaluation

HyDE retrieval, parent document retriever, structured output, and RAG evaluation

Criteria

Without context

With context

HyDE hypothetical generation

100%

HyDE retrieval uses hypothetical doc

100%

HyDE state fields

100%

HyDE graph ordering

100%

Parent doc child chunk size

Parent doc parent chunk size

RAGResponse Pydantic model

30%

100%

Structured output binding

100%

Faithfulness metric

100%

Retrieval quality metrics

100%

Answer quality metrics

25%

100%

Repository: Dicklesworthstone/pi_agent_rust
Commit: 47823e3

Evaluated: 29 days ago
Agent: Claude Code
Model: Claude Sonnet 4.6

Table of Contents

Internal Documentation Q&A Assistant Technical Support Knowledge Base Search Scientific Literature Q&A with Quality Evaluation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.