skills.rest
Harmitx7
Harmitx7COMMUNITY
Explore all skills by this author →
WebsiteGitHub

llm-engineering

Design RAG pipelines and select embeddings for LLM systems.

3|1|Updated Feb 20, 2026
Search Tags:#prompt engineering#llm#rag#vector database#ai systems#llmops
One-click install
npx skills add https://github.com/Harmitx7/tribunal-kit --skill llm-engineering
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: llm-engineering
Source: https://github.com/Harmitx7/tribunal-kit/tree/main/.agent/skills/llm-engineering
Command: npx skills add https://github.com/Harmitx7/tribunal-kit --skill llm-engineering

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill provides the foundational knowledge and best practices for building robust, reliable, and efficient AI systems that leverage Large Language Models (LLMs), preventing common pitfalls like hallucinations and cost overruns.

Core Features & Use Cases

  • LLM Architecture Patterns: Understand when to use simple prompts, RAG, fine-tuning, or agent loops.
  • RAG Pipeline Design: Learn optimal chunking, embedding, and retrieval strategies.
  • Prompt Engineering: Master message structuring and few-shot examples for predictable outputs.
  • Evaluation & LLMOps: Implement strategies for measuring performance and managing production concerns like cost and reliability.
  • Use Case: When developing a new AI-powered customer support chatbot that needs to answer questions based on your company's documentation, this Skill guides you through designing an effective RAG pipeline and prompts.

Quick Start

Use the llm-engineering skill to design a RAG pipeline for a new AI feature.

Frequently Asked Questions about llm-engineering

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

FAQPage Schema
How do I design a RAG pipeline to prevent LLM hallucinations?▼

To design a RAG pipeline that prevents hallucinations, apply optimal chunking strategies, select appropriate embedding models, and use reliable vector stores to ground LLM responses in factual documentation.

What is the best way to manage LLM costs and reliability in production?▼

Managing LLM costs and reliability in production requires implementing LLMOps strategies, which involve measuring performance metrics, controlling inference expenses, and applying robust architecture patterns to ensure consistent AI system behavior.

When should I use RAG instead of fine-tuning for my AI system?▼

You should use RAG instead of fine-tuning when your AI system needs to answer questions based on frequently updated documentation, whereas fine-tuning is better for adapting the model's tone or specific behavioral patterns.

How do I structure prompts for predictable LLM outputs?▼

To structure prompts for predictable LLM outputs, use message structuring techniques and few-shot examples, which guide the model's reasoning and enforce consistent formatting across multi-turn conversational agents.

Can I build a multi-turn conversational agent using LLM architecture patterns?▼

Yes, you can build multi-turn conversational agents by applying specific LLM architecture patterns, using prompt engineering to maintain context and ensure reliable interactions throughout the conversation.

What are the limitations of using simple prompts for semantic search?▼

Simple prompts lack external knowledge context for semantic search, limiting their effectiveness; overcoming this requires RAG pipeline design with vector databases to retrieve and embed relevant information accurately.

Related Skills

Explore More Software Engineering Skills→
amurataamurata
6

rag-implementation

Implement RAG systems with vector databases and embedding models.

Community
Advanced
reggiechan74reggiechan74
5

administration

Configure OpenTelemetry exporters to monitor Claude Code usage and costs.

Community
Advanced
rukikotoorukikotoo
6

rag

Query the econometrics RAG knowledge base to retrieve relevant document chunks.

Community
Intermediate
Clay-HHKClay-HHK
2

model-merging

Merge fine-tuned models into a single model using YAML configurations and mergekit.

Community
Advanced