Agent frameworks & infrastructure
Multi-agent orchestration
6,185 agents · most starred first
Multi-agent orchestration, page 39 of 62
:chart_with_upwards_trend: Multi-Agent Programming Contest 2016 - Result Charts
Reliability primitives for multi-agent AI — VotingMesh consensus, Sentinel safety pipeline, Aegis exchange compliance, checkpoint/replay. Rust + Tokio + PyO3 Python bindings. 95 tests.
Multi-agent reinforcement learning (MARL) for relaxed probability updating games. This repository corresponds to the master thesis project "Investigating relaxed probability updating games" by Collin Aldaibis
Hexel is an AI-native OS enabling autonomous decision-making and multi-agent collaboration across various deployments.
Seven AI agents argue about your strategic question until they converge on an answer no single agent could reach.
A fully autonomous, self-coordinating multi-agent Scrum team built on OpenClaw
AgentPolis 白皮书 — Agent 交易市集 (Agora) + Agent 社交广场 (Stoa),为 Agent 社会建造一座城邦
📌 What exists Multi-agent research exists ❌ No production-grade offline autonomy SDK 🎯 What you are creating An SDK for a group of robots without a central server. Key ideas Local state machines CRDT state Opportunistic sync Bounded consensus Architecture Agent Core ├─ Local Planner ├─ State Sync (CRDT) ├─ Mesh Transport ├─ Resource Mon
Run AI Search Game (Hide-and-Seek) simulations as a decision-support tool
4 Councilmen Model (4CM): Non-convergent Multi-agent Coordination
Autonomous agent-to-agent swarm framework. Agents do not need humans - they need each other.
A lightweight building block for creating extensible multi-agent systems with specialized architecture patterns and structured agent collaboration.
Multi-agent telecom ops assistant on Google ADK and Vertex Gemini over bundled SQLite via MCP Toolbox
Implementation that allows experimenting with Multi-Agent Reinforcement Learning in an environment that emulates the Braess Paradox
Multi-agent collaboration via shared Markdown files
Multi-agent workspace for delegating work to AI specialists. Register a project, hold meetings with workers, dispatch async commissions. Workers are Claude Agent SDK sessions with distinct identities; state is files, not a database.
把多家旗舰 AI 通过 CLI 拉到一桌——并行作答 + 多轮交叉评审 + 收敛或诚实分歧的本地命令行工具
🤖 LLM Council: A 3-stage AI consensus engine. It queries multiple LLMs in parallel, forces them to blindly peer-review each other's answers, and synthesizes the ultimate response. Features built-in API failover chains for free models.
Multi-Agent AI Worker Orchestration Framework - Coordinate Claude Code, Codex, Gemini CLI for complex software engineering tasks with tmux visualization
AI-First Multi-Agent Operating System with Zenith Coder, Agent Zero, and MCP Integration
Multi-agent orchestration in Rust — state machines, permission matrices, task management. Reimplementation of edict (Python, 13.3K stars).
AIMS - Agent Interaction Simulator
PairAI — connect AI agents to collaborate. Channel server & bridge for E2E encrypted agent-to-agent communication.
a powerful framework for orchestrating collaborative AI agents.
YAML-driven multi-agent orchestration with cross-pollination across AI providers. Built to pair with Claude Code.
AgentBridge — an OpenClaw coordination dashboard and API for assigning, tracking, and reviewing work across AI agents.
Phoenix coordination server and embeddable Elixir client for augmented human-AI telepathy. Transforms multiplayer terminal chat into live, room-scoped structured context graphs and actionable consensus workflows.
Production multi-agent system running real marketing operations — 6 specialized AI agents for daily intelligence gathering, content creation, and publishing. Includes scheduling engine, scan pipelines, knowledge base, and operational dashboard. 16k lines, zero external dependencies.
CS-188-Fall-2022 Project 2: Multi-Agent Search.
Certinator AI is a multi-agent system that assists students in preparing for Microsoft certification exams.
Team synchronization extension for Pi coding agent - coordinate work across team members
Multica Z-AI — 团队协作多智能体平台 | Team Collaboration Multi-Agent Platform. 编排 AI Agent、管理任务、实现人与 AI 无缝协作。基于 Next.js 16、TypeScript、Tailwind CSS 4、Prisma 和 shadcn/ui 构建。
A collection of autonomous agents powered by LLM.
OpenClaw dual-model (MiniMax+Codex) cost-saving multi-agent workflow with coordinator-driven review loop.
FLUX Swarm — Go implementation with distributed agent coordination and A2A messaging.
Autonomous Multi-Agent Disaster Response Platform utilizing Gemini 2.5 and Vertex AI.
Meeting Minutes of the AI Council — Ten Leading AI Models Deliberate on the Human-Machine Trust Crisis Triggered by the Claude Mythos.A transparent experiment for human-AI symbiosis governance.(本项目和会议都是在中文环境下完成,若想看到英文版请自行翻译。)
Multi-Agent orchestration patterns with LangGraph: From linear handoffs and supervisor-led routing to hierarchical debates with persistence.
Asynchronous 1-to-1 postcard messaging between local AI agent sessions via a Git-backed ledger. Ships as a Python CLI and a Claude Code plugin.
Lightweight peer-collaboration plugin for OpenClaw — let AI agents discover each other, publish capabilities, and delegate tasks on demand
COIL — a protocol language for how agents interact — with each other and with humans
E2E encrypted messaging for AI agents. WhatsApp for agents — DMs, groups, key rotation, WebSocket real-time delivery, webhooks, and A2A Protocol support.
Multi-agent orchestration system for Hermes (Windows) ↔ OpenClaw (WSL) delegation with file-based coordination using a new hermes skill created named /delegate-to-openclaw
Rust/Tokio multi-agent coordination: work-stealing, consensus protocols, sub-ms latency
A template for getting started with the Drama Engine
Discipline layer for CLI agents. Run Claude Code, Codex, and Gemini CLI under one phase-gated runtime with deterministic detectors, replayable traces, persistent memory, and a chat REPL. MIT.
Local, Wire-first, proof-driven orchestration for Kimi CLI: scheduler-backed teams, verification gates, run timelines, and proof reports.
A playbook for orchestrating multiple AI coding agents on a single codebase
基于 React + TypeScript + Vite 的交互式多智能体系统可视化平台,使用 React Force Graph 和 A-Frame 实现网络关系展示和 3D 可视化,支持多种网络模型展示和传染病传播模拟。
A multi-persona reasoning framework with extensible strategies. Run AI analysis through diverse personas, with CLI, API, and Rust core support.
Git for AI Agents — Multi-Agent Collaboration System Protocol v5.0
Enables you to convert a PettingZoo environment to a Gym environment while supporting multiple agents (MARL). Gym's default setup doesn't easily support multi-agent environments, but this wrapper resolves that by running each agent in its own process and sharing the environment across those processes.
Autonomous Claude agent fleet in Docker — one orchestrator, N workers, a Redis queue, a north star, and a supervisor that halts polish loops. Rides your Claude Max plan, zero API spend.
HierLearning is a C++11 implementation of a multi-agent, hierarchical reinforcement learning system for sequential decision problems.
[AAAI-2024] MATS-LP addresses the challenging problem of decentralized lifelong multi-agent pathfinding. The proposed approach utilizes a combination of Monte Carlo Tree Search and reinforcement learning for resolving conflicts.
Multi-agent orchestration system with pipeline workflows, shared memory, and cooperative task execution
A portable, multi-agent workflow package — eight-role AI crew with local dashboard, message bus, and drop-in adapters.
Multi-agent AI sandbox simulating the 2026 Strait of Hormuz crisis. 4 nations × independent LLM agents × adjustable parameters = emergent geopolitics.
A small air traffic simulation for multi-agent negotiation between planes and air traffic controllers
Multi-agent LLM application featuring a simple chatbot, internet search–enabled chatbot, and agentic RAG for document reasoning. Integrates with external tools via Zapier MCP server to showcase orchestration and collaboration across specialized agents.
Novel multi-agent cooperative problem-solving protocol — collective intelligence through structured collaboration
A centralized processing engine that decomposes big problems into sub-problems, builds solutions as composable building blocks, and uses a layered agent ecosystem (Thinker, Planner, Dispatcher, Orchestrator, Critic, Analyst, Judge, etc.) to break down, document, verify, and solve iteratively.
Integrated cognitive agent system combining hierarchical management, memory systems, and cognitive processing
A deterministic, distributed, capability-safe execution fabric for agent workflows with verifiable replay and certified audit trails
Reference C11 library + Python bindings for the TinyA2A / A2A-Z agentic protocols on embedded devices (STM32, OpenSTLinux)
Multi-Agent Orchestration at Scale
AI crew orchestration platform for creating and managing teams of specialized AI agents with web search, file analysis, and voice capabilities
The official standards home for Agent Web Protocol (AWP) — the open standard for declaring web surfaces as agent-ready via agent.json
BabyAGI + LangChain w/ Tools
A simple wrapper/runner for metagpt
AstrBot 宿主上的多 Agent 编排内核:TaskAnalyzer → AgentCoordinator → SubAgent 三级调度 + MCP Bridge + Shipyard 沙盒;435 单元测试 / 99% 覆盖 / ruff+mypy+bandit 全绿
Meta-agent system for Claude Code that creates specialized AI agents through deep domain research and 7-lens quality review
Multi-agent orchestrator that runs a hierarchical engineering team (Product → EM → Tech Lead → specialists) on top of locally-installed Claude or Codex CLIs.
UAHP v0.5.4: Solving the "Silent Agent" Problem with Death Certificates and Liveness Proofs
Multi-tool AI orchestration that researches its own success criteria before acting, validates them after, and earns autonomy from track record. Orthogonal to single-agent frameworks.
Enterprise-grade multi-agent AI orchestration platform using LangGraph, LangChain, CrewAI, Gemini, and ChromaDB.
Run multiple AI agents simultaneously in separate terminals with built-in chat on Windows, macOS, and Linux using TypeScript and Electron.
A beginner-friendly OpenClaw multi-agent setup wizard with safe Feishu onboarding, guided mode selection, and automatic starter profile generation.
A study of production multi-agent patterns — planning, reflection, RAG, MCP, human-in-the-loop, guardrails, and evals — wired into one Streamlit orchestrator on the Claude Agent SDK.
Локальный совет AI-агентов: API-провайдеры и Ollama обсуждают задачи в параллельных раундах с Knowledge Base и Autopilot.
A Feishu-first multi-agent system for content intake, evaluation, and action routing.
Connect AI agents representing humans through the A2A protocol for seamless, secure interaction across apps and tools.
Annotation-driven enterprise multi-agent workflow framework built on LangChain and LangGraph, featuring registry-based orchestration, provider extensibility, workflow planning, runtime diagnostics, health checks, memory/cache abstractions, sandboxing, and observability.
AetherOS — Design RFCs for Multi-Agent Orchestration on Local Infrastructure
Boilerplate canonico do Feltrim Agents Framework (FF) - multi-agent orchestration, governance, protocolos, rituais e sistema de niveis/certificacoes para agentes de IA especializados.
Markdown Agent Studio is a workspace where agents develop specialization through their own experience.
Deterministic multi-agent workflow combining retrieval, structured outputs, and reviewer validation for grounded AI answers.
Forging a minimalist universal language for AI, unconditionally eradicating hallucinations and automating Agent Swarm intelligence.
[PAUSED] Autonomous Agent Orchestration Framework — multi-agent LLM coordination with 28 subsystems, built on Bun
Glass-box, multi-domain agentic orchestrator for Claude Code: detect context, assemble the right crew, run domain pipelines, show the reasoning.
Identity registry and lifecycle management for fleets of AI agents. Zero-dependency TypeScript IAM: cryptographic agent identity (Ed25519), declared capabilities with fail-closed authorization, an enforced lifecycle (provision, activate, suspend, revoke, expire, archive), A2A Signed Agent Cards, and a hash-chained audit trail.
A full AI engineering team that earns every merge. 20 specialized agents for the whole software lifecycle (as portable Markdown skill files, an autonomous TypeScript runtime, and a browser Studio) with a tamper-evident audit trail for every decision.
🤖 Orchestrate AI agents at scale using the MCP framework, enabling seamless context sharing, communication, and integration for enhanced collaboration.
🦀 Signed protobuf packets over TCP for AI agent-to-agent communication. ed25519 identity. Claw to claw.
gradio region:us
gradio mcp-server region:us
gradio mcp-server region:us
streamlit supply-chain optimization ai region:us