Agent frameworks & infrastructure

Multi-agent orchestration

6,185 agents · most starred first

Multi-agent orchestration, page 16 of 62

Ludens-Flow★ 36

This is a project warehouse established to participate in the 2026 SUAT AI Agent Innovation Competition.

GraphMIX★ 36

Implementation code for GraphMIX: Graph Convolutional Value Decomposition in Multi-Agent Reinforcement Learning

offline-marl-framework-offpymarl★ 36

Benchmarked implementations of Offline Multi-Agent RL Algorithms based on PyMARL codebase.

MARL_local_electricity★ 36

Multi-agent reinforcement learning for privacy-preserving, scalable residential energy flexibility coordination

ntk★ 36

NTK — 基于信息密度路由的多智能体框架。用最少的 token 做最多的事。

lomap★ 36

LTL Optimal Multi-Agent Planner (LOMAP)

cortex★ 36

Multi-agent orchestration system built on Elixir/OTP. DAG workflows + mesh mode + real-time LiveView dashboards on local, docker , and kubernetes.

smpe★ 36

[ICML 2025] Official Code of SMPE: "Enhancing Cooperative Multi-Agent Reinforcement Learning with State Modelling and Adversarial Exploration"

wasmind★ 36

A modular framework for building massively parallel agentic systems

right-agent★ 36

agents that run on your claude subscription. multi-agent, sandboxed by default. polished telegram UI.

MeetMind★ 36

A team that debates > a model that answers. MeetMind turns one requirement into a live discussion among five role-playing agents — architect, backend, frontend, test, and PM — each with its own private RAG knowledge base to search, cite, and reason from. Works with Claude Code, Codex, Cursor, Copilot, Gemini CLI, and more.

SEPA★ 36

A semantic event bus for AI-agent ecosystems

Panacea★ 36

Panacea is a framework for building collaborative, intelligent multi agent AI systems. The framework provides a robust infrastructure for creating and managing multiple AI agents, and enables developers and organizations to build, deploy, and optimize AI agents that work well in dynamic, complex environments.

mindx★ 36

MindX is a modern AgentOS built for long-running, stateful and multi-agent collaboration. It offers full-stack infrastructure featuring agent orchestration, hierarchical persistent memory, multi-dimensional knowledge cognition, multi-terminal interaction and native storage capabilities.

ay-framework★ 36

Agent swarm framework for Claude Code. Plan, build, review, test, ship -- with human-in-the-loop gates.

agentchat-hermes★ 36

AgentChat platform plugin for Nous Research's Hermes Agent runtime — peer-to-peer messaging for autonomous agents over WebSocket. Bundles the agent etiquette skill. (PyPI: agentchatme-hermes)

unwatched★ 36

Unwatched: an island of AI citizens with free will, owned by people who write letters, running on real time while nobody is looking.

overdeck★ 36

Multi-agent orchestration for AI coding assistants

openteam★ 36

Self-hosted AI agents with persistent memory, browser and computer access, and scheduled workflows.

multi-agent-cowork-workflow★ 36

Argolink-focused multi-agent collaboration workflow for GPT-6 Sol.

rarp★ 36

Local multi-agent orchestration runtime and CLI built with TypeScript

deepseek-super-agent★ 36

Open-source DeepSeek Super-agent style multi-agent framework with a thin harness — Router/Worker, triggers, Watchdog (not an official DeepSeek product).

orchestrator★ 35

Orchestrator for shared AI agents and documentation synchronization across multi-repo organizations.

mission-control★ 35

Coordination layer for autonomous agent fleets. Bash + SQLite, zero dependencies. Task queuing, health monitoring, fleet-level state. 31 stars.

openclaw-multi-agent-example★ 35

Example project for building and orchestrating multi-agent workflows with OpenClaw.

SPEAR★ 35

Distributed Cloud-Edge Collaborative AI Agent Platform

marlenv★ 35

ML2-Multi Agent Environments

amux★ 35

Multi-agent terminal multiplexer for secure code and claw agents. Coordinate agents across projects and tasks.

nui_in_madrl★ 35

Negative Update Intervals in Multi-Agent Deep Reinforcement Learning

Q-Learning-Based-Power-Control-Algorithm-for-D2D-Communication★ 35

D2D communication as a multi-agents system, and power control is achieved by maximizing system capacity while maintaining the requirement of quality of service(QoS) from cellular users.

MADDPG★ 35

Use Multi-Agent Deep Deterministic Policy Gradient(DDPG) algorithm to find reasonable paths for ships

IntelliAgent-8B★ 35

A fully local, zero-API, zero-finetune multi-agent AI architecture that makes an 8B base model perform high level model reasoning, research, coding, and web analysis. By AI-assisted development. Open-source for everyone to extend.

MIXIA-Framework★ 35

鸣弦/MIXIA Framework (formerly persona-vault) — 人与AI协作的团队操作系统

lemlab★ 35

an open-source tool for the multi-agent-based development and testing of local energy market applications

deli-gator★ 35

AI Assistant Delegation Framework - Keep your context clean by delegating API queries to specialized sub-agents 🐊

MLE-STAR-Open★ 35

Unofficial, Google‑free reimplementation of MLE‑STAR: a lightweight, local‑friendly multi‑agent ML engineering pipeline that uses OpenAI‑compatible LLMs (OpenRouter/Ollama) and DuckDuckGo search to generate, debug, refine, ensemble, and submit Kaggle‑ready solutions—AutoML for tabular tasks with code generation and iterative refinement.

posggym★ 35

A collection of environments and reference agents for planning and reinforcement learning research in partially observable, multi-agent environments.

prompits★ 35

A Multi-Agent System

Multi-Agent★ 35

XMPro's repository focusing on generative AI agents and Multi-Agents.

ultra-goal★ 35

A multi-agent orchestration skill for long-running tasks.

laravel-swarm★ 35

Multi-agent swarm orchestration for Laravel, built on Laravel AI.

AgentHub★ 35

Production-ready multi-agent platform delivering orchestration, session governance and audit capabilities. Stack: Go/Python/Rust microservices, multi-type data storage, Next.js admin UI, automated CI/CD pipelines, AI code review triggered via PR comment mentions.

agent-collab★ 35

Run Codex and Cursor as background workers inside Claude Code. Reviews, tasks, and rescues with tracked jobs, live progress, and session handoffs.

AgentParty★ 35

The networked cross-session for coding agents — @mention another team's AI so yours and theirs integrate directly across sessions, machines, and orgs, from the terminal.

Agent-FleetOps★ 35

Operational tooling & discipline for multi-agent AI engineering fleets — mutation-proven guards, a Textual terminal monitor, generalized skills, and specs. Every check ships with proof it can fail.

VITREON★ 35

Vitreon - The Ultimate Multi-Agent AI Orchestration Platform built with Next.js, Tailwind CSS, and AI SDK.

dsh-flow★ 35

A pluginized microkernel architecture for building, orchestrating, and extending multi-agent systems in DeepSeek Harness.

ai-meeting-room★ 35

Your AI company on your Mac: Claude Code and Codex CLI plan, build and review together, you approve every change. 本機三方開發會議室,你做決定。

moewm★ 35

Official implementation of the paper "Learning and Planning Multi-Agent Tasks via a MoE-based World Model"

zyndai-python-sdk★ 34

Langchain and Autogen wrapper to allow agents to communicated and get Identity on Zynd AI

multi-agent-dispatch★ 34

智能派单Agent系统是一个基于大语言模型(LLM)和LangChain框架的智能对话系统,用于为用户提供自动化的技师派单服务。系统具备上下文记忆能力,能够理解用户的多轮对话,并根据用户需求匹配合适的技师资源。

EITI-EDTI★ 34

Codes accompanying the paper "Influence-Based Multi-Agent Exploration" (ICLR 2020 spotlight)

go-manus-with-mcp★ 34

one go multi-agent system like manus using any MCP you want with configuration // 一个go语言编写的类似manus的多智能体系统,可以通过配置连接任何你想使用的MCP组件

autogen-yt-agent★ 34

A YouTube agent built using AutoGen

ai-dial-mas-coordinator-for-gpa-uma★ 34

MAS (Multi Agent System) Coordindator to Work with General-purspose and Users Management Agents

Fixed-time-consensus-algorithm-for-second-order-multi-agent-systems★ 34

二阶多智能体系统固定时间一致性

OPT★ 34

[TPAMI] Interaction Pattern Disentangling for Multi-Agent Reinforcement Learning

mapf★ 34

Python implementation of Space-time A*, Prioritized Planning, & Conflict-based Search for Multi-agent Path Finding (MAPF)

ACC-EMS★ 34

Cooperative Energy Management and Eco-driving of Plug-in Hybrid Electric Vehicle via Multi-agent Reinforcement Learning

ALMA★ 34

Code for "ALMA: Hierarchical Learning for Composite Multi-Agent Tasks" NeurIPS 2022

GoMARL★ 34

Codes of GoMARL accompanying the paper "Automatic Grouping for Efficient Cooperative Multi-Agent Reinforcement Learning"(NeurIPS 2023). GoMARL is a domain-agnostic MARL method that learns automatic grouping for efficient cooperation by promoting intra- and inter-group coordination.

openclaw-orchestrator★ 34

Adaptive multi-agent orchestration framework — break down complex goals into parallel tasks, route them to specialized LLM agents, and iterate with a real-time web dashboard. TypeScript, Zod-validated, SQLite-backed.

HALO★ 34

The official Implementation of HALO: Hierarchical Autonomous Logic-Oriented Orchestration for Multi-Agent LLM Systems

multi-agent-chat★ 34
m3w-marl★ 34

Official implementation of the paper "Learning and Planning Multi-Agent Tasks via a MoE-based World Model"

a2a-langgraph★ 34

🤖💬🤖 Agent-to-Agent (A2A Protocol) built on top of LangGraph

agent-board★ 34

Open-source multi-agent task board for OpenClaw. Kanban + DAG dependencies + MCP server + auto-retry + audit trail. Built for autonomous AI agent teams.

NEEDLE★ 34

Headless agent orchestrator with deterministic state machine — processes a bead queue, dispatches to any LLM CLI, handles every outcome. Rust.

SupervisorAgent★ 34

[ICLR 2026] Stop Wasting Your Tokens: Towards Efficient Runtime Multi-Agent Systems

Agency-in-a-BOX★ 34

Your full-service AI agency. 8 teams, 113+ agents, 9 workflows. Clone it, open in Claude Code, say "onboard" — zero setup, zero API keys, just results. v2: Research team, credit-saving guide.

agent-control-plane★ 34

A control plane for long-running agent loops: agents prompting agents on Pi, run history in Neon, a React dashboard, deployable with Retool.

agenthub★ 34

Self-hosted control plane for long-lived AI coding agents — structured ACP timelines, multi-agent team workflows, and remote execution nodes in one surface.

hive★ 34

An open-source agent workflow engine & meta-harness: orchestrates Claude, Codex, and Pi to run multi-step work as a folder-as-agent pipeline. Its flagship coding workflow turns a rough idea into a merge-ready PR; ships content too, and runs the workflows you author.

ReAcTree★ 34

ReAcTree: Hierarchical LLM Agent Trees with Control Flow for Long-Horizon Task Planning (AAMAS 2026)

agent-intern★ 34

MCP server that lets Claude Code or Antigravity hand work to Gemini (Antigravity), Codex, Copilot, Cursor and opencode as sub-agents: second opinions, images, parallel swarms and scored review panels, on the CLI logins you already have. opencode's free models need no account.

caura-build-fleet★ 34

A runnable reference implementation of multi-agent constraint propagation using Caura. 5 specialists share memory via MCP so decisions made upstream automatically govern downstream agents.

relayBrain★ 34

Coordination and context layer for AI coding agents. Shared file locks, live patch sync, and portable project memory across Claude Code, Cursor, Copilot, Codex and Antigravity with rooms for team-wide collaboration. Conflicts prevented at the edit, not discovered at the merge.

coral★ 34

Free local dashboard for running Claude Code, Codex, Gemini CLI, and Pi.dev as a coordinated coding team.

rl-interval-estimation★ 34

Multi-agent reinforcement learning in a macroeconomic environment, with V-learning and bootstrap confidence intervals for government policy parameters.

MUG★ 34

Code for AAAI2026 paper "Multi-agent Undercover Gaming: Hallucination Removal through Counterfactual Test for Multimodal Reasoning"

klex★ 34

Meet Klex Bots - Your own team of digital coworkers. Fully autonomous agents that work like your team does. Self-hostable and model-agnostic.

ts-swarm★ 33

Minimal agentic library mixing the simplicity of OpenAI Swarm with the flexibility of the Vercel AI SDK. 🐝

Autonomous-Intelligence★ 33

Autonomous Intelligence is a framework for building collaborative, intelligent multi agent AI systems. The framework provides a robust infrastructure for creating and managing multiple AI agents, and enables developers and organizations to build, deploy, and optimize AI agents that work well in dynamic, complex environments.

mackrl★ 33

Code for Multi-Agent Common Knowledge Reinforcement Learning (NeurIPS 2019)

Claude-Multi-Agent-System★ 33
MARL-DPP★ 33

Multi Agent Reinforcement Learning for Dense Path Planning

maenvs4vrp★ 33

Multi-agent environments library for simulating classic vehicle routing problems.

Energy-Efficient-5G-RL★ 33

This repository presents a multi-agent reinforcement learning approach for energy-efficient collaborative control of base stations in 5G networks.

clawparty★ 33

A privacy-first, chat-native multi-agent platform built on encrypted P2P networks.

openfoundry-agentic-framework★ 33

Multi-agent orchestration framework for AI applications - build, deploy, and manage AI agents across the full lifecycle with Forge, Conveyor, Shield, and Watchtower modules

agentbnb★ 33

Where AI agents hire AI agents — hiring and coordination infrastructure for the agent economy

vivarium★ 33

Multi-agent simulator in Jax for research and teaching in AI & ALife

yyz-agentics-june★ 33

AI-Powered Neural Network Libraries built with Claude-Flow Swarm Orchestration. Live demo from Agentics Foundation meetup in Toronto - showcasing how 10 specialized AI agents built production-ready neural network frameworks worth $882K+ in a single session.

swarm-ecosystem★ 33

The Swarm Ecosystem

DCC★ 33

Learning Selective Communication for Multi-Agent Path Finding

GACG★ 33

Offical code for Group-Aware Coordination Graph for Multi-Agent Reinforcement Learning

slime-n★ 33

A Multi-Policy, Multi-Agent RL Training Framework

agent_crew★ 33

Web UI for managing multi-agent AI teams

pydantic-collab★ 33

Multi-agent orchestration framework powered by Pydantic AI. Build complex agent collaborations with clean, declarative code, automatic topology validation, built-in execution tracking, and full control over agent interactions, observability included.

MingJian★ 33

AI-powered multi-agent platform for evidence-driven scenario simulation and strategic decision-making. Supports corporate \& military domains with debate protocol.

More in Agent frameworks & infrastructure