What LLM providers are supported by this skill?

The skill includes implementation patterns for both OpenAI and Anthropic APIs, covering message creation, streaming, and configuration settings.

How does the skill handle API rate limits and errors?

It includes a 'safeLLMCall' utility that implements automatic retries with exponential backoff specifically designed to handle 429 rate limit errors.

What prompt engineering techniques are included?

It covers structured system prompts, few-shot examples for classification, and chain-of-thought reasoning to improve model accuracy on complex tasks.

Does this skill help with RAG implementation?

Yes, it provides comprehensive code patterns for the entire RAG pipeline, including document chunking strategies, embedding storage, and vector similarity search.

Does it provide guidance on cost and token management?

Yes, the skill includes best practices for token tracking, caching common queries, and utilizing cost-effective models for simpler tasks.

LLM Application Development

Name: LLM Application Development
Author: januarfonti

byjanuarfonti

Data Science & ML

Builds robust AI-powered applications using advanced prompt engineering, RAG patterns, and multi-provider LLM integrations.

About

The LLM Application Development skill equips Claude with specialized patterns for architecting production-ready AI features. It provides standardized implementation strategies for structured prompting, few-shot learning, and chain-of-thought reasoning, alongside optimized code for OpenAI and Anthropic API integrations. By incorporating RAG (Retrieval-Augmented Generation) workflows—including document chunking and vector storage with pgvector—this skill streamlines the creation of intelligent chatbots, automated agents, and context-aware systems while ensuring best practices in error handling, cost management, and streaming response delivery.

Key Features

Complete RAG pipeline implementation with chunking and vector search
Real-time streaming response management for low-latency UX
0 GitHub stars
Resilient error handling with exponential backoff and retries
Advanced prompt engineering patterns including CoT and Few-shot learning
Multi-provider API integration for OpenAI and Anthropic SDKs

Use Cases

Building streaming AI interfaces for real-time user interaction
Implementing automated text classification and sentiment analysis tools
Developing context-aware AI chatbots using internal knowledge bases

About

Key Features

Complete RAG pipeline implementation with chunking and vector search
Real-time streaming response management for low-latency UX
0 GitHub stars
Resilient error handling with exponential backoff and retries
Advanced prompt engineering patterns including CoT and Few-shot learning
Multi-provider API integration for OpenAI and Anthropic SDKs

Use Cases

Building streaming AI interfaces for real-time user interaction
Implementing automated text classification and sentiment analysis tools
Developing context-aware AI chatbots using internal knowledge bases