react-pattern

A Collection of Pydantic Models to Abstract IRL

🤖 Native iOS implementation of the ReAct (Reasoning + Acting) AI agent pattern with SwiftUI. Watch AI think, act, and solve problems step-by-step using OpenAI's GPT models.

TinyAgent brings the core ideas of the original Python library into a lean, zero-dependency TypeScript package. It turns plain functions into LLM-driven tools and lets you chain them together through a simple ReAct-style loop.

Design patterns and reference implementations for autonomous AI agent systems. ReAct, tool-use, memory management, fallback, and multi-agent orchestration.

ARA — Agent Runtime Architecture: Java 21 framework for building autonomous AI agents and multi-agent systems. LLM integration, tool calling, deterministic I/O contracts, RAG and human-in-the-loop. On Maven Central (io.github.xmor).

Terminal AI agent with ReAct pattern - execute shell commands, run Python, analyze images. Supports Gemini, OpenAI, Ollama.

🤖 An AI-powered DevOps agent implementing ReAct architecture to automate mobile release pipelines (Crash Check -> Perf -> Build -> Audit). Built with LangGraph & Ollama.

A coding agent that edits real repos with workspace tools (read/write/search/run), produces JSON traces with token accounting, and renders runs on a live dashboard.

Loop 工程学习手册：让 AI Agent 自己跑起来。系统梳理 ReAct、Reflexion 到 Claude Code /loop 的演进与工程实践。

A step-by-step hands-on guide to building a tool-using AI Agent from scratch in Python with the Google Gemini API. Zero heavy frameworks, just core concepts.

Production-ready template for complex AI agents using LangGraph and LangChain. Implements RAG, ReAct, routing, parallelization, and orchestrator patterns with FastAPI backend and PostgreSQL checkpoints

零框架从零构建生产级 AI Agent — 四周手写 ReAct / MCP / LangGraph / AutoGen / Multi-Agent 协作

日报 AI 自动化框架：AI Agent + Workflow 混合架构，端到端实现从 Excel 面板入职到每日日报自动生成与飞书通知

LLM weather agent built on the MCP protocol, combining Groq, LangChain and multi-API orchestration to answer natural-language queries with real-time data, forecasts and historical context.

AI agent systems skill for Claude Code — ReAct, Plan-and-Execute, reflection loops, Tree of Thoughts, LATS, MCP integration, LangChain/LangGraph/CrewAI/AutoGen/Claude Agent SDK, multi-agent orchestration, evaluation. By TheGEOLab.net

A comprehensive multi-agent AI system built with LangGraph and Google Gemini. Features specialized agents (Coder, Enhancer, Researcher, Validator), RAG implementation, ReAct pattern, and supervisor orchestration. Includes data analysis, document drafting, and conversational agents with quality validation.

AI-powered article generator for Dev.to and Medium

A guarded NL-to-SQL reasoning agent that learns schema and domain context, then answers analytical questions using iterative, evidence-backed SQL.

A self-reflective LLM agent with tools, memory, and reasoning built using LangChain + ReAct + Reflexion. Modular FastAPI backend + Streamlit UI.

AI research agent using ReAct pattern (Reason + Act). Built with Next.js + z-ai-web-dev-sdk. Free tools: DuckDuckGo search, r.jina.ai URL reader, Wikipedia API.