How does this skill help with agent tool selection?

It provides standardized patterns for ReAct and OpenAI Function calling, ensuring tool descriptions and prompt structures are optimized for accurate agent decision-making.

Can I use this for production RAG pipelines?

Yes, it includes blueprints for document loading, intelligent text splitting, and vector store integration to build high-performance retrieval-augmented generation systems.

Does it support different memory types for chat history?

It covers a range of memory patterns including simple buffers, summarized context for long conversations, sliding windows, and semantic vector-store retrieval.

How do I handle LLM costs and monitoring with this skill?

The skill includes comprehensive guidance on implementing the LangChain callback system to track token usage, latency, and detailed request/response logging.

LangChain Architecture & LLM Agents

Name: LangChain Architecture & LLM Agents
Author: Tahir-yamin

byTahir-yamin

•

Data Science & ML

Designs and implements sophisticated LLM applications using LangChain's modular framework for agents, memory, and multi-step workflows.

LangChain Architecture is a specialized Claude Code skill designed to help developers build production-grade LLM applications. It provides comprehensive patterns for orchestrating autonomous agents, managing conversation state via various memory types, and constructing complex multi-step chains. Whether you're implementing Retrieval Augmented Generation (RAG), integrating external tools through function calling, or optimizing document processing pipelines, this skill offers the architectural guidance and code patterns needed to create scalable and efficient AI-driven systems within the LangChain ecosystem.

Key Features

01Advanced memory management for short-term and long-term conversation context

02Modular chain composition for complex, sequential LLM workflows

03RAG pipeline implementation with vector store and document loader patterns

04Multi-agent system design including ReAct and OpenAI Function patterns

05Observability and performance monitoring through custom callback handlers

063 GitHub stars

Use Cases

01Building autonomous AI agents with specialized tool access and decision-making logic

02Developing RAG-based systems for semantically querying internal documentation

03Creating complex, multi-step data transformation and analysis pipelines using sequential chains

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add tahir-yamin/dev-engineering-playbook langchain-architecture

For use in Claude.ai and ChatGPT

Download Skill