piyushagni5/langgraph-ai

Implements production-ready agentic RAG systems with adaptive routing and self-correction, MCP (Model Context Protocol) server/client patterns, and human-in-the-loop orchestration workflows. Built around LangGraph's state machine architecture, it provides practical patterns including tool-calling agents, multi-step reasoning chains, parallelization, and evaluator-optimizer loops. Covers integration across retrieval systems, orchestrator-worker topologies, and streaming HTTP transports for real-world AI application deployment.

100 stars.