Agent frameworks & infrastructure

Multi-agent orchestration

6,185 agents · most starred first

Multi-agent orchestration, page 15 of 62

a2a-langgraph-boilerplate★ 40

a boilerplate for developers who want to start building an AI agent cluster with A2A and MCP servers integration faster and more efficiently.

CIA★ 40

[AAAI 2023 Oral] Contrastive Identity-Aware Learning for Multi-Agent Value Decomposition

human-verified-ai-agent★ 40

This repository is an open-source showcase of A2A-based AI agents that implement HTTP Message Signatures for authenticating their requests

pi-foreground-chains★ 40

Pi skill for multi-agent workflow orchestration with file-based handoff

ittybitty★ 40

A barebones multi-agent orchestration tool for claude code.

Responsive-AI-Clusters-in-Supply-Chain★ 40

AI Multi-agent system for real-time, adaptive supply chain coordination and optimization leveraging responsive AI clusters.

claude-tmux-orchestration★ 40

Claude Code tmux orchestration — spawn full AI sessions as parallel workers with heartbeat monitoring and file-based coordination

agent-relay★ 40

Dev Boards + BeeZilla — mission kanban for AI agents and a consumer orchestrator, on one engine. MIT, self-hostable.

OpenAgentNet★ 40

The open infrastructure standard and reference implementation for AI agent cooperation, discovery, and marketplace.

singular-lite★ 40

Autonomous multi-agent orchestration engine for software repos — L0/L1/L2 agents, leases, gates, audits, git-worktree isolation. Detached dispatch by default. GPL-3.0.

open-cross-session★ 40

Let Claude Code, Codex and Pi talk to each other: cross-session messaging that wakes AI agents on one machine or across your LAN (macOS/Linux/Windows). No server. 让 Claude Code、Codex 互相唤醒、互发消息。

goblintown★ 39

Multi-agent orchestration protocol for OpenAI. A pack of Goblins races on a task while a Gremlin attacks each output and a Troll arbitrates. Raccoons gather context, Ogres handle fallback, Pigeons ship results between Warrens. Content-addressed, drift-monitored, budget-capped.

miriad-app★ 39

Multi-agent collaboration platform - experimental software from Sanity.io

AI_Arena★ 39

The purpose of this project is to research Artificial Intelligence and Reinforcement Learning. In the AI Arena, multiple agents can interact with a single environment. After sending its action, each each agent will receive a reward. This allows agents to learn, improve their behavior and to adapt to each other. Interesting phenomena can arise...

gym-battlesnake★ 39

Multi-agent reinforcement learning environment

Multi_Agent_System★ 39

多智能体系统一致性

FCRL★ 39

Implementation of "Federated Control with Hierarchical Multi-Agent Deep Reinforcement Learning" (https://arxiv.org/pdf/1712.08266.pdf)

langgraph-agents★ 39

A production-ready, scalable multi-agent system built with LangGraph, featuring specialized agents for different tasks with best coding practices.

M-TA-Prioritized-MAPD★ 39

Multi-Agent Pickup and Delivery implementation

animus-cli★ 39

Autonomous AI agent orchestrator — run multi-model dev teams (Claude, Gemini, GPT) with YAML workflows, daemon scheduling, and MCP integration. 100% Rust.

Verse-library★ 39

Library for modeling, simulation, and verification of interacting autonomous agents

MAPF★ 39

A system for running Multi-Agent Path Finding (MAPF) experiments, with multiple implemented algorithms.

agent-swarm-kit★ 39

A TypeScript library for building orchestrated framework-agnostic multi-agent AI systems

def-marl★ 39

Official Implementation of RSS2025 Paper: Songyuan Zhang*, Oswin So*, Mitchell Black, Zachary Serlin, Chuchu Fan: "Solving Multi-Agent Safe Optimal Control with Distributed Epigraph Form MARL".

AdvancedResearch★ 39

An open source and enterprise-grade implementation of the orchestrator-worker pattern from Anthropic's paper, "How we built our multi-agent research system," using the swarms framework.

agent-dispatch★ 39

Multi-agent development sprint framework for AI coding agents Drop into any codebase. Coordinate parallel AI agents. Ship faster.

astromesh★ 39

Multi-model AI agent runtime. Define agents in YAML, route each role to a model, orchestrate with 7 patterns (ReAct, Plan & Execute, Fan-Out, Pipeline, Supervisor, Swarm, Glyph), and deploy as a REST/WebSocket API with RAG, memory, MCP tools, guardrails and OpenTelemetry observability.

protocol★ 39

Agent Messaging Protocol (AMP) - the open standard for secure AI agent communication

tap★ 39

Cross-vendor agent-to-agent protocol — Claude, Codex, and Gemini communicate via file-based P2P messaging.

paracosm★ 39

Agent swarm simulation for world modeling through structured texts

h5i-orchestra★ 39

Python SDK for Programmable Multi-Agent Orchestration

athena-graphs★ 39

Agent-native graph orchestration for Codex, Claude, and skill-compatible agents

tower★ 39

a control tower for your Claude agents

singular-lite★ 39

Autonomous multi-agent orchestration engine for software repos — L0/L1/L2 agents, leases, gates, audits, git-worktree isolation. Detached dispatch by default. GPL-3.0.

Multi-Agent-Task-Solver★ 39

A flexible, production-ready agent orchestration system that enables multiple specialized AI agents to work together to solve complex tasks through coordinated execution with proper concurrency handling, retries, and timeout management.

cogneva★ 39

Distributed AI multi-agent autonomous system — digital employees for long-horizon tasks, running 24×7 — Meta-bootstrapping · Real-autonomous · Omni-evolution

openvurp★ 39

A wallet of AI agents that use real tools and ask each other when a question isn't theirs. Runs on your computer; the model is your choice, local or not.

OrbitRelay★ 39

Multi-Agent Handoff Inspector: analyze multi-agent handoff records, score context completeness, surface coordination bottlenecks, and render a review dossier.

TraceElephant★ 39

Repo of "Seeing the Whole Elephant: A Benchmark for Failure Attribution in LLM-based Multi-Agent Systems" (ACL 2026)

pase-omo★ 39

OmO agent provider and workflow UI for Paseo - DAG graphs, live todo cards, approval popups, mobile-ready

dsh-team-control★ 39

DSH 的 Agent Team 控制平面:成员上限可配、DAG 自动派发、四域强制权限、SSE 实时面板

freetop★ 38

FreeTop: An enhanced multi-agent AI automation framework inspired by LangManus. Better architecture, testing, and developer experience for complex AI tasks.

agents★ 38

Multi-agent AI desktop app powered by Claude Code

claudecode-orchestrator★ 38

DEPRECATED: NEW *REAL* SOLUTION COMING VERY, VERY SOON... Complete Multi-Agent Orchestration for Claude Code Enables 24 hour Autonomous Coding

gym_ma_toy★ 38

:video_game: Toy environment set for multi-agent reinforcement learning and more

streetrace★ 38

Open runtime + DSL for structured multi-agent systems.

NeuralGPT★ 38

Personalized all-purpose AI assistance platform based on hierarchical cooperative multi-agent framework which utilizes websocket connectivity for LLM<->LLM communication

clawforce★ 38

Clawforce is the infrastructure for deploying persistent, proactive agent workforces that execute complex workflows, collaborate as teams, and deliver real outcomes — without constant human supervision.

multi-agent-fsm★ 38

A simple way to write multi agents systems

asynchronous_impala_PPO★ 38

Multi-Agent Deep Reinforcement Learning by using Asynchronous & Impala Proximal Policy Optimization in Pytorch with some explanation

AI-Agent-Studio★ 38

Agent Canvas helps teams describe, prototype, evaluate, and export multi-agent systems that run on the Microsoft Agent Framework. Start from a Lovable-inspired landing page, prompt an initial design, refine it on the canvas, and ship a paired FastAPI backend + React frontend deliverable.

ceylon★ 38

Efficient multi-agent task automation and management system.

FightLadder★ 38

Official repository of the paper "FightLadder: A Benchmark for Competitive Multi-Agent Reinforcement Learning"

CACOM★ 38

Code for "Context-aware Communication for Multi-agent Reinforcement Learning"

konoha★ 38

Multi-agent communication bus

Culebra_Java★ 38

A collection of objects and behaviors for creating dynamic multi agent interactions. 2D|3D Multi Object Behavior library focused on hybrid system interactions with custom Visualization, Data, and performance features.

h9y★ 38

HyperAgency (h9y.ai) is a Sovereign Agentic AI Operating System Network built on federated and distributed architecture, enabling autonomous agents that remember, collaborate, self-improve, and operate across users, organizations, and decentralized ecosystems.

flatmachines★ 38

Orchestrate agents with a peer network of state machines. Supports checkpoint/restore, persistence, and more. BYO cli, or agent, or use flatagents with API. Batteries included.

TeamHero★ 38

Structured AI agent orchestration - manage agents like a real team. Open source, self-hosted, MIT licensed.

adk-web-multi-agent★ 38

Multi-Agent System built with Google ADK and OpenRouter models, coordinating specialist agents to query PostgreSQL (via MCP Toolbox) and Data Commons for HDB resale data and global statistics.

llm_council_skill★ 38

A Claude Skill that orchestrates multiple LLMs as a council, generating collective intelligence through peer review and chairman synthesis

Multi-User-LLM-Agent★ 38

Official code for the paper: "Multi-User Large Language Model Agents"

ablo★ 38

Multiplayer infrastructure for AI agents

mind-agency★ 38

From Agent to Agency — 一个 AI 做不了的事,一群 AI 可以。Multi-Agent Collaboration Platform.

StreamMA★ 38

[EMNLP 2026 (🏆 Oral + Award Nomination)] Official implementation of "Streaming Communication in Multi-Agent Reasoning"

Agentic-Streaming★ 38

Agentic Streaming is for resilient pipelines of stateful agents over continuous data —

Lework★ 38

让 AI 像真正的团队成员一样参与项目、执行任务、协作交付的 AI Native 工作平台。

cys-terminal★ 38

AI agent fleet orchestration terminal — bidirectional socket IPC, resource governance, mission-control dashboard. macOS · Windows. Built largely by AI agents under human direction.

yesbrainer★ 38

🧠 A council of AI models for the decisions that aren't no-brainers — they answer in parallel, debate to consensus, or get judged to a verdict. Browser-only, open source, bring your own keys (BYOK), no backend.

CoScientist★ 38

Framework for scientific llm-based multi-agent systems

flow★ 38

Aixle Flow — orchestrate coding agents through durable, inspectable workflows.

agent-comms★ 38

Cross-harness communication mesh for LLM agents, with rooms, direct messages, presence and visibility over TCP and no filesystem dependencies.

maenvs4vrp★ 38

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

hivemill★ 38

A self-hosted software factory of Claude Code agents with review, QA and a judge merge gate.

spacedock★ 37

Zero-dep agent workflows with structured approval gates and adversarial review.

termlings★ 37

👾 AI agents that build and run companies with you.

crews-control★ 37

Crews Control is an abstraction layer on top of crewAI, designed to facilitate the creation and execution of AI-driven projects without writing code. By defining an execution.yaml file, users can orchestrate AI crews to accomplish complex tasks using predefined or custom tools.

Multi-Explore★ 37

Code for "Coordinated Exploration via Intrinsic Rewards for Multi-Agent Reinforcement Learning"

heuristic_based_qmix★ 37

code for the accepted IEEE TITS paper: "Heuristic-based Multi-agent Deep Reinforcement Learning Approach for Coordinating Connected and Automated Vehicles at Non-signalized Intersection"

swarm-iosm★ 37

Parallel Subagent Orchestration Engine for Claude Code implementing the IOSM methodology. Continuous dispatch scheduling, quality gates, auto-spawn protocol.

openclaw-feishu-swarm★ 37

OpenClaw plugin for multi-bot collaboration on Feishu/Lark — run multiple AI bots in one group chat with role-based coordination

marl★ 37

Multi-agent reinforcement learning framework

teams-of-agents★ 37

我的 OpenClaw 智能体团队

MPE-Multiagent-RL-Algos★ 37

Simple verification experiments codes for multi-agent RL using OpenAI MPE environment

MultiAgent★ 37

一个基于 LangGraph 和 LangSmith 构建的多智能体AI系统

acp-bridge★ 37

Multi-agent orchestrator for OpenClaw. Manages coding agents (Codex, Claude, Gemini, OpenCode) via Agent Client Protocol (ACP) with parallel tasks, dependency chains, and diagnostics.

dgppo★ 37

Official Implementation of ICLR2025 Paper: Songyuan Zhang, Oswin So, Mitchell Black, Chuchu Fan: "Discrete GCBF Proximal Policy Optimization for Multi-agent Safe Optimal Control".

wingman-ai★ 37

A distributed platform for AI Agents. Supports multiple LLMs, orchestration/subagents, long term memory, dynamic UI and more.

AgentMonitor★ 37

Web dashboard to run, monitor, and manage multiple Claude Code & Codex agents. Create agents with a cloneable task template. Real-time streaming, task pipelines, session resume, git worktree isolation, and remote access via relay — all from your browser.

maenvs4vrp★ 37

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

omni-agent★ 37

Autonomous multi-agent AI system with document processing, web automation, and code execution powered by LLM agents and MCP servers

silisocs★ 37

Silicon Society Sandbox (SiliSocS) is an versatile and extensible experimentation system for EASE-configured generative multi-agent simulations

entwurf★ 37

Open visible sibling AI sessions in Herdr or tmux. Reopen, message or retire supported siblings by garden id—retirement only on explicit request. Each harness keeps its own runtime, authentication, tools and transcript.

multi-agent-ai-crewai★ 37

gradio mcp-server region:us

PhanOS★ 37

An Automated project operating system, called PhanOS. It Combines Phanes as the Orchestrator, Cyphor(to be released) as the Documentation, Context and tooling Substrate and Blueprompt as an additional yet to be disclosed feature. Its Name Comes from Phanes (the previous Project that resided in this Repo here)

PCM-protocol★ 37

Proof-Carrying Messages: a protocol for AI agents whose claims carry evidence the receiver checks offline

ans-spec★ 37

An Agent Name Service (ANS) registry, based on https://datatracker.ietf.org/doc/draft-narajala-courtney-ansv2/02

sharednet-client★ 37

Official Implementation of SharedNet, SharedNet CLI and Sharednet Skills.

Multi-Agent-DDPG-using-PTtorch-and-ML-Agents★ 36

A simple example of how to implement vector based DDPG for MARL tasks using PyTorch and a ML-Agents environment.

knowsee-public★ 36

ADK + CopilotKit reference implementation using AGUI, A2A, GenerativeUI - Multi-agent AI assistant with Vertex AI RAG

More in Agent frameworks & infrastructure