Can I use this to build autonomous agents?

Yes, it contains detailed logic for building agents that use reasoning and tools, including the ReAct loop and function calling patterns.

Is this skill suitable for complex multi-step tasks?

Absolutely. It specifically features the 'Plan-and-Execute' and 'Agent Team' patterns designed for decomposing and solving complex, multi-stage tasks.

How does this help with RAG implementation?

It provides code patterns for various chunking strategies, vector database selection (Pinecone, PGVector, etc.), and advanced retrieval techniques like hybrid search.

What architectural patterns are included in this skill?

The skill includes patterns for RAG pipelines, ReAct agents, function calling architectures, plan-and-execute frameworks, and multi-agent collaboration teams.

Does it support LLMOps and monitoring?

Yes, the skill covers production-ready patterns for prompt versioning, observability, and monitoring essential for maintaining LLM applications in production.

LLM Application Patterns

Name: LLM Application Patterns
Author: claudiodearaujo

byclaudiodearaujo

•

データサイエンスとML

Provides production-ready architectural blueprints and code implementation guides for building advanced, scalable LLM-powered applications.

This skill serves as a comprehensive reference for designing and implementing sophisticated AI systems within the Claude Code environment. It offers detailed architectural patterns for Retrieval-Augmented Generation (RAG) pipelines, diverse agentic frameworks such as ReAct and Plan-and-Execute, and complex multi-agent collaboration strategies. By leveraging these industry-standard implementation patterns, developers can move beyond basic chat interfaces to build robust AI applications with integrated document ingestion, hybrid search, prompt management, and LLMOps monitoring.

主な機能

011 GitHub stars

02Multiple agent design patterns including ReAct, Function Calling, and Plan-and-Execute.

03Prompt IDE patterns for variable management, versioning, and A/B testing.

04Comprehensive RAG pipeline architectures covering chunking, embedding, and storage strategies.

05Advanced retrieval strategies such as Hybrid Search, Multi-query, and Contextual Compression.

06Multi-agent collaboration frameworks for coordinating specialized AI roles.

ユースケース

01Implementing a multi-agent workflow where specialized agents (researcher, writer, critic) collaborate on content creation.

02Developing autonomous agents capable of planning complex tasks and using external tools via function calling.

03Building a knowledge-based enterprise assistant using hybrid vector search and RAG.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add claudiodearaujo/sistema-de-narra-o-de-livro llm-app-patterns

For use in Claude.ai and ChatGPT

Download Skill