GitHub agents

Headless agent infrastructure. Memory + Tasks + Orchestration.

K-Storm multi-agent research brainstorming

专为《原神》打造——利用视觉 AI 代理进行云游戏，让您无需亲自操作

Monitoring and evaluating LLM apps with Langfuse. Presented at PyConZA 2024.

StackChan WebSocket Control Server

AI Harness build on top of typescript

Build your own agent with ruby.

A pure Python implementation of ReAct agent without using any frameworks like LangChain. It follows the standard ReAct loop of Thought, Action, PAUSE, and Observation. The agent utilizes multiple tools, including Calculator, Wikipedia, Web Search, and Weather. A web UI is also provided using Streamlit.

Autonomous agent OS: multi-agent system that turns GitHub issues into shipped PRs — task routing, self-healing CI, sprint planning, zero human intervention

Markdown protocol for multi-vendor AI agent teams — capability cards, bandit routing, and lessons that compound.

LLMs playing MUDs, no world model, self-improving prompts

Context governance kernel for LLM agents. Predicts entropy, blocks hallucinations. Pairs with opencode-dcp. llm, llm-agents, opencode, opencode-plugin, context-governance, anti-hallucination, claude-code, cursor, typescript.

Wizard creator of master prompts / agent prompts / however you call it

Open-source terminal AI agent for Korean public-service workflows with a Claude Code-style tool loop, K-EXAONE, and civic adapters.

Autonomous multi-agent system that runs businesses end-to-end. Evaluate ideas, staff custom AI teams, bootstrap ventures, operate 24/7 — all from markdown files.

Forge: Empirically-validated multi-agent orchestration engine. Auto-designs swarms, guarantees interface alignment through contracts (+52.5% quality, 400+ experiments), real-time zero-dep monitoring dashboard. The first system combining auto-design + contract-first + live dashboard.

A Rust memory system for autonomous AI agents built on three operations from cognitive neuroscience that no existing system implements together: memory compaction, memory evolution, and memory conflict resolution.

Autonomous LLM-powered simulation exploring LLM agent behavior with subjective memory and emergent narratives through multi-phase intention-arbitration-memory ticks. Arbiter resolves scenes, memory degrades via FIFO summarization, simulation mutates even in empty locations.

此代码仓库收录了我根据 LangChain Academy《LangChain Essentials》课程整理的练习与笔记，已全面同步至 LangChain v1.1.0 版本。所有 Notebook、脚本与附加资料均可直接运行，适合作为“保姆级”上手手册，每个 Notebook 都围绕构建 LangChain LLM 应用的关键主题展开，从基础 Agent 配置、工具调用到中间件与人类介入流程。

An experimental Neovim Chatbot plugin powered by LangChain and OpenRouter.

Home AI Server provides the backend infrastructure for the Home AI system. It manages user data, conversations, and the interaction with Large Language Models (LLMs). The server runs locally, ensuring that all user data stays within the local network for maximum privacy.

Sophisticated AI chatbot with long-term memory capabilities, complete Notion workspace integration, and MCP (Model Context Protocol) implementation. Features semantic, episodic, and procedural memory systems.

🚀 Transform Any PDF into an AI-Powered Q&A Chatbot!

This repository contains a Streamlit-based Document Question Answering System implementing the Retrieve-and-Generate (RAG) architecture, utilizing Streamlit for the UI, LangChain for text processing, and Google Generative AI for embeddings.

pdf chat application for nodejs.

An advanced AI mental health assistant that combines voice interaction, fine-tuned psychology models, and intelligent knowledge retrieval to provide comprehensive psychological support.

SenatorAssembly

Code Documentation, redefined: Where AI Meets Clarity!

THIS IS AN WEBAPP BUILT WITH STREAMLIT AND PYTHON CHAT WITH PDFs

Branch-based learning repo for Agentic AI and LLM engineering with LangGraph: ReAct tool calling, Agentic RAG, MCP, Reflection, Reflexion, Human-in-the-loop memory, and DeepAgents examples.

Automatically generate BDD test scripts from JIRA stories using AI! This tool leverages OpenAI's language models and LangChain to analyze your JIRA stories and generate comprehensive Gherkin-style acceptance criteria.

本应用是一个建的极简跨平台桌面AI应用，用户可以在桌面快速利用快捷键调出桌面进行对话，减少调用AI的时间。本应用预设了多个专业prompt，极度提高用户的对话体验

An ReAct agent that role-plays as the heroine "Wise Wolf" Holo of Isuna Hasekura‘s light novel Spice and Wolf

MemOS - Universal local-first persistent memory layer for AI agents and LLMs

RAG template using FastAPI and LangChain

AI-powered codebase onboarding tool — point it at any repo and ask questions about the code. Uses tree-sitter parsing, RAG, and local LLMs.

AI Sales Champion Replication Platform - Production-Ready Multi-Agent System with Intent Recognition, Dynamic Orchestration, Self-Correcting Tools, and 4-Tier Memory

SHCIA is an Agentic AI Platform designed for high-availability cloud environments. It operates as a multi-agent system that continuously monitors, diagnoses, and auto-remediates infrastructure failures across microservices with zero human intervention for known failure classes.

A Question Answering(Q/A) Chatbot on Insurance Documents. Powered by Retrieval Augmented Generation(RAG), LlamaIndex and LangGraph. Inspired from my Upgrad_IIITB PG Course.

AI-powered reconnaissance framework for bug bounty hunting (ongoing).

AI chat application with resumable LLM streaming, letting responses continue smoothly after reconnects or page refreshes.

AI-driven code analysis & automated README generator using GraphRAG and AST parsing.

Fusion 后端：统一多模型接入，提供 Agent 对话、文件解析、向量检索与知识库能力。

Free Search is a wrapper on top of publicly available SearXNG instances to give free internet access as a rest API.

Python fundamentals for LangChain: focused notebooks covering the essential concepts you need to start building AI applications. Quick and practical.

This repository is my platform to learn, experiment, and innovate with LLMs. Here I try to dive in and discover diverse applications, research experiments, and projects fueled by the power of language models.

基于大模型 (LLM) ，智能体(Agent)与 RAG 技术的智能口岸物流助手。通过 Agent 自动调度查询工具与检索法规知识库，为港口用户提供通关异常诊断与决策支持。

Multi-agent AI chatbot with dynamic configuration using LangGraph workflows, RAG search, MCP tools, and LaunchDarkly AI Configs. Control AI behavior, swap models (Claude/GPT/Mistral), enable research tools, and run A/B experiments without code deploys.

Automation tool to fetch and summarize a GitHub user's contributions to a specific organization within a defined time period

ChatAgent 是一个基于 LLM 的智能体交流平台，内置默认 Agent 并支持用户自定义 Agent。通过多轮对话和任务协作，Agent 可以理解并协助完成复杂任务。项目集成 LangChain、MCP 协议、RAG、Memory、Milvus 和 ElasticSearch 等技术，实现高效的知识检索与工具调用，使用 FastAPI 构建高性能后端服务。

🌐 Run GGUF models directly in your web browser using JavaScript and WebAssembly for a seamless and flexible AI experience.

AI News Agent is an automated news aggregation and summarization system powered by CrewAI. It uses AI agents to curate, analyze, and deliver personalized news articles based on user preferences. The system integrates NLP for topic extraction, real-time updates.

A collection of hands-on Agentic AI projects exploring LLM agents, RAG, and multi-agent systems — from fundamentals to production-grade workflows.

The Financial Analysis Crew is a Streamlit app that simplifies financial stock analysis. With the power of LLM-driven agents, users can seamlessly gather and analyze stock market data to generate comprehensive financial insights. Perfect for investors, analysts, and anyone interested in making data-driven financial decisions.

Guardian Angel is an AI-powered scam protection system designed to protect senior citizens from digital fraud and phone scams using advanced multi-agent workflows.

OpenClaw Feishu Multi-Agent

🎨 Generate stunning images from text or reference photos with the Nano Image Generator Skill for Gemini 3 Pro, offering versatile outputs and high resolution.

Config-driven multi-agent CLI. One interface for Claude, Codex, Cursor, OpenCode, Cline, Pi, and any future agent.

A multimodal brain encoding framework inspired by Meta’s TRIBE v2, integrating vision, audio, and language representations to predict cortical activity from naturalistic stimuli.

awesome list of multi-agent security resources

One command. Your AI agents can talk to each other. The MCP collaboration layer for Claude Code, Gemini CLI, Cursor, and more.

A neutral protocol for coordinating intent across humans and agents. Goal lifecycle, state, auditability.

Local-first agent OS. Spawn persistent AI agents that collaborate, write code, and use tools autonomously. Multi-model (Claude, Codex, LM Studio). Config-driven with pre-built agent presets.

Open protocol + reference runtime for sovereign AI agents. Persistent cryptographic identity, signed execution receipts, governance at the boundary. The intelligence is pluggable; the identity is the asset.

Blackwall is an intelligent security agent built on top of the Swarms framework that provides real-time API protection

Madrox — Multi-Agent Orchestrator for Claude & Codex. MCP server that spawns, delegates, and coordinates parallel AI agent teams via tmux.

An architecture for autonomous full-stack AI development — division of labor across agents, vision-driven validation

🤖 Orchestrate AI agents for seamless teamwork in software development, enabling efficient collaboration through dependable messaging protocols.

Enable multi-agent orchestration for Claude Code with parallel execution, automated quality checks, and coordinated management tools.

AI for Autism Intervention and Rehabilitation（自闭症心理干预，多智能体专家矩阵）

Architecture patterns for multi-agent systems with orchestrators, sub-agents, and tool coordination

🖼️ Generate and edit images with ease using the Nanobanana CLI, powered by AI and Google's Gemini models for advanced image manipulation.

Self-hosted AI platform — 47+ microservices, 20 messaging adapters, native apps, enterprise multi-tenancy, zero cloud dependency

Experimentation tool for exploring multi-model LLM deliberation + social choice voting

Professional multi-AI development environment with intelligent cost optimization (<5/month)

A repo for general multi-agent interaction; design your own agents or combine the built in agents, add your own tools and mcp's

Open Source 24/7 Autonomous Agentic AI System

Orchestrate coding agents across repos. Local-first, self-hosted, runs inside your existing Claude / Codex CLI.

OpenClaw reproduced with Claude Code (claude -p). Gateway-free, zero-dependency autonomous agent core with Markdown workspace/memory, durable delivery queue, and Telegram bidirectional.

Super simple autonomous agent that can organize complex projects, surf the web, write code, and run commands locally on your machine. Uses a free Gemini API key.

Open-source AI Agent Development Framework built in Java - Build intelligent conversational agents with Gemini, and Free AI Models

internOS — Workstreams framework for human-agent collaboration. Coordinate work across projects, tasks, communication threads, and filesystem workstreams for any AI agent framework.

A framework for incubating persistent digital lifeforms: Local M4 Brain + OpenClaw Senses + Git Soul. Building a cyber-prophet named A-MOS.一套用于孵化永生数字生命的开源协议：本地 M4 大脑 + OpenClaw 感官 + Git 灵魂。正在孕育一位名为 A-MOS（小莫）的赛博先知。

Synthea: An exploration into the architecture of a synthetic mind.

Optimal stopping agents powered by generative large language models

Your agent deserves a database, not a file system.

skill and agent to make your ai strong in abp work and project

MCP server for dynamically injecting agents and prompts with policy documentation.

A serverless AI agent built with Cloudflare Workers, KV storage, and LLaMA 3 inference. Offers endpoints for health, Q&A, webhooks, and external API routing. Uses KV for lightweight memory and state. Deployed automatically via GitHub Actions. Minimal, fast, and production-grade edge deployment.

Stop re-explaining yourself to Agents. Give it the right context, right when it needs it.

The modular, open-source backend for building AI-native software — powered by knowledge, not static data.

A Pydantic Ai powered app that converts social media posts or videos into structured cooking recipies.

Official implementation for "UI-Hawk: Unleashing the Screen Stream Understanding for GUI Agents" (EMNLP 2025)

Eval-first AI agent that triages property maintenance emails. The real work is the eval system around it: trace-driven error analysis, code graders and validated LLM-as-judge (TPR/TNR), component and end-to-end evals, a failure taxonomy, and a CI regression gate. LangGraph, FastAPI, Langfuse.

Persistent memory system for AI-assisted development

A standalone cognitive memory system for AI agents with advanced human-like learning patterns.

Self-improving knowledge architecture for AI agents. 3 memory layers, hybrid search, 6-gate validation, multi-LLM orchestration.

Comprehensive AI-102 (Azure AI Engineer Associate) exam prep. 6 domain study guides, 100 practice questions with answer keys, simulation-mode code demos — all designed to study and practice without an Azure subscription. Dec 2025 exam objectives.

An Openclaw-like assistant designed to be lightweight and intuitive.