Does this skill support both OpenAI and Anthropic?

Yes, it includes production-ready implementation patterns for both OpenAI and Anthropic SDKs, including support for streaming and structured message formats.

Does it provide guidance on vector databases?

Yes, it specifically includes patterns for using Supabase with pgvector for storing and searching document embeddings.

How does the skill handle API rate limiting?

It includes a 'safeLLMCall' utility that implements exponential backoff and configurable retry logic to handle 429 rate-limit errors and other transient failures.

Can I build a RAG system using this skill?

Absolutely. The skill provides full patterns for document chunking, embedding generation, and vector database queries to build Retrieval-Augmented Generation systems.

LLM Application Development

Name: LLM Application Development
Author: julianromli

byjulianromli

•

101

データサイエンスとML

Builds robust AI-powered applications through advanced prompt engineering, RAG implementation, and resilient API integration patterns.

概要

This skill empowers developers to build sophisticated applications powered by Large Language Models (LLMs) by providing a comprehensive library of design patterns and best practices. It covers essential domains including structured prompt engineering (few-shot, Chain of Thought), multi-provider API integrations for OpenAI and Anthropic, and full Retrieval-Augmented Generation (RAG) pipelines. With built-in logic for document chunking, vector database management, and resilient error handling like exponential backoff, this skill transforms Claude into an expert architect for AI-driven features and autonomous agents.

主な機能

101 GitHub stars
Resilient Error Handling: Robust wrappers for managing API rate limits, retries, and fallback mechanisms.
RAG Pipeline Orchestration: Comprehensive logic for document chunking, embeddings, and vector database retrieval.
Multi-Provider SDK Support: Implementation templates for OpenAI and Anthropic integration including streaming responses.
Vector Database Integration: Patterns for storing and searching embeddings using solutions like Supabase pgvector.
Advanced Prompt Engineering: Patterns for structured prompts, few-shot learning, and Chain of Thought reasoning.

ユースケース

Building a context-aware knowledge base chatbot using RAG patterns.
Implementing reliable AI automation with built-in retry logic and cost management.
Developing specialized classification tools using few-shot prompt engineering techniques.

概要

主な機能

101 GitHub stars
Resilient Error Handling: Robust wrappers for managing API rate limits, retries, and fallback mechanisms.
RAG Pipeline Orchestration: Comprehensive logic for document chunking, embeddings, and vector database retrieval.
Multi-Provider SDK Support: Implementation templates for OpenAI and Anthropic integration including streaming responses.
Vector Database Integration: Patterns for storing and searching embeddings using solutions like Supabase pgvector.
Advanced Prompt Engineering: Patterns for structured prompts, few-shot learning, and Chain of Thought reasoning.

ユースケース

Building a context-aware knowledge base chatbot using RAG patterns.
Implementing reliable AI automation with built-in retry logic and cost management.
Developing specialized classification tools using few-shot prompt engineering techniques.