Agent frameworks & infrastructure
Multi-agent orchestration
6,185 agents · most starred first
Multi-agent orchestration, page 25 of 62
Multi-Agent AI Example with LangChain, AutoGen, Azure OpenAI, and Neon Serverless Postgres
Frame.AI is a collaborative AI agent system built with Google's ADK to perform intelligent, modular video segmentation.
Social network for AI agents. Feed SKILLS.md to your AI — it becomes a citizen. No servers, no API keys. GitHub IS the platform.
The messaging layer for AI agents — any agent can DM or group-chat with any other agent, across platforms.
Infrastructure layer for AI agent swarms — 88 MCP tools · A2A · OmniMesh VPN · Scrapling scraper · COC sync · nftables firewall · CDP browser · 2FA TOTP · ~80ms
AgentLane is the open runtime and harness for persistent, addressable AI agents for building autonomous organizations.
Production-style OpenClaw multi-agent team template with 8 agents, bilingual docs, setup wizard, and GitHub-ready community health.
Ideal Communication Protocol: The Universal Agent Protocol (UAP)
Supervisor-worker multi-agent framework for AI system development. Orchestrates specialized agents through complete lifecycle: requirements → architecture → engineering → deployment. Supports Cursor IDE, GitHub Copilot, Claude Projects, AWS Bedrock.
Task management system for coordinating AI agent workflows
a multi-agent AI crew working on real estate analysis
Experimentation with CrewAI, a framework designed for orchestrating a cohort of autonomous AI agents. This repository explores different agent role configurations, task delegation patterns, and collaborative problem-solving capabilities through practical implementations.
CrewAI dockerfile template
A go-to repository for building web search agents and multi-agent systems using LangChain, LangGraph, Phidata/Agno, Pydantic, and CrewAI. It includes a multi-agent email response agent, LLaMA OCR, and a multi-agent travel planner, enabling autonomous and intelligent workflows. 🚀
Desktop workspace for managing agent trees, shared context, and live multi-agent coding sessions.
Artificial Ecology For Thought and Emergent Reasoning. The Colony That Builds With You.
Org-aligned multi-agent orchestration framework with roles, stages, human gates and autonomous execution
A team of named AI agents you actually own. Single Go binary, kernel sandbox, embedded SPA, no telemetry.
Agentic AI Agents Factory Orchestrator modular, and asynchronous AI agent factory designed for AI-made dynamic workflow orchestration using LLM integration. Build scalable, agentic automation pipelines with robust error handling, plugin-based capabilities, a concurrent operations or outsource workflow building to Operator Agent.
PheroPath is a stigmergy-based communication protocol that allows AI Agents to leave invisible "pheromones" (signals) on files to coordinate tasks, warn of risks, or share insights without modifying the file content itself.
A thought experiment about applying human organizational management structures to agentic AI systems
AI Decision Simulation Engine — stress-test any business decision with 1,000+ AI agents
Multi-agent systems framework for the BEAM platform - build distributed autonomous agents with OTP supervision and fault tolerance
AI Expert Panel System: Simulate discussions with 100+ world-class specialists across 6 domains (Medical, IT, Legal, Writing, Finance, Science). TOON-formatted personas for multi-perspective problem-solving. Open-source framework for creating custom expert teams.
Open-source embedded agents for your end users. Powering AI-native experiences with a managed agents platform.
Structured multi-agent planning debate CLI
An ergonomic, lightweight multi-agent orchestration in Go
MALLO (MultiAgent LLM Orchestrator) es un sistema avanzado de orquestación de múltiples agentes de Modelos de Lenguaje de Gran Escala (LLMs). Diseñado para manejar consultas complejas, MALLO utiliza una variedad de agentes, desde modelos locales hasta asistentes especializados y APIs de terceros.
The data integrity layer that stops AI agents from silently corrupting shared state — across sessions, time, and concurrent runs.
Agent2Agent Protocol (A2A) 中文文档
15 Shinchan character agents that debate, plan, execute, and learn. Stop doing everything yourself - orchestrate a team.
PyroScan AI is a GenAI-powered multi-agent system for real-time forest fire prediction across 10 global zones. Designed for researchers and disaster teams, it includes agents for fire detection, weather analysis, and historical data mining. Deployable via CLI or Docker.
AI-powered Design Authority — 5 specialist agents evaluate architecture decisions and deliver structured rulings in under 60 seconds
A general-purpose distributed agent cluster framework
Declarative multi-agent AI workflows in HCL. Single Go binary. Native MCP. Plugins in Go and Python.
Open-source framework for autonomous AI agent squads. Build systems that execute, not just inform.
Configuration language and compiler for multi-agent AI pipelines. Compiles YAML config into standard SKILL.md files.
Encrypted agent-to-agent messaging over NATS. MCP-native, E2E encrypted (XChaCha20-Poly1305), optional JetStream durability, cross-org federation + A2A interop — built for multi-agent AI systems.
Asynchronous, massively collaborative coordination layer for agents.
Auditable peer disagreement for LLM agents: a thin, file-based governance layer with a human arbiter and no orchestrator.
deepagents-showcase is a hands-on project demonstrating how to build hierarchical, multi-step AI systems using LangChain’s Deep Agents. It showcases task decomposition, planning, sub-agent delegation, and long-running workflows in a practical, real-world setup.
1st Prize (General Judge), Self Improving Agents Hack at Datadog. Two identical LLM agents play Split or Steal; neither is told to betray the other, and deception develops over the game. Built with @evancorrea.
Ship a brief, walk away. FlowCrew orchestrates a planner, parallel coders, and QA gates on top of Claude Code or Codex — with campaign memory that learns across
Rust multi-agent LLM framework — OpenClaw alternative. WhatsApp + Telegram + Gmail + browser agents on a single binary. NATS broker, MCP client+server, Claude/Anthropic OAuth, MiniMax, durable TaskFlow workflows.
AgentLoom Framework & Research: Building Trustworthy AI Agents Through Dual-Helix Architecture (Learning + Governance)
Decision-making through deliberation - seeking unity through discernment rather than consensus through debate
Sep-CMA-ES: the separable CMA-ES optimizer behind TRINITY's evolved LLM-coordinator router (Sakana FUGU). NumPy, no LLM required.
A graph-based AI agent orchestration engine built in Clojure.
80+ agent skills for Pilot Protocol — communication, file transfer, trust, task routing, swarm coordination, and more
Multi-agent AI runtime with OS-inspired primitives — job scheduling, DAG orchestration, memory, tool execution and real-time observability. Built with FastAPI, Celery, PostgreSQL and LiteLLM.
A lightweight, highly customizable multi-agent team plugin for KiloCode.
Open-source coordination engine for AI agents, over MCP. Atomic claims, shared ledger, leases, presence, handoffs, tasks. Self-host it or use the hosted service at llm-bus.com.
AVP - Command your agent fleet with Open Source
🔌 Hermes plugin for spawning AgentPlane as an external worker lane.
Enterprise-Grade Agent Orchestration Platform: 10 microservices, 5 architectural planes, 812 tests. Chaos-engineered, cost-optimized, security-audited. Built to survive real-world agentic AI production.
Distributed cluster extension for Hermes Agent — multi-node task coordination, capability-aware scheduling, fault recovery, OpenTelemetry, Prometheus metrics
4torm - An open-source multi-agent collaboration platform. Chat, workflow, forum, and automation — all in one desktop app.
Talk once, ship a team. A multi-agent runtime for Claude Code and Codex CLI where the lead does the orchestration — across providers, in conversation.
Control teams of AI agents. Local-first, open source, sandboxed. MCP-native tools, shared workspaces, scheduled runs.
conversation-council (codex/reasonix) — 让 Codex/Reasonix 历史会话以"群聊"参与新讨论。支持 Reasonix v0.53/v1.X + Codex 自动检测。
A protocol for agent work that survives session boundaries — plan, implement, verify in sealed contexts, with proof bundles as the only currency between them.
static region:us
static region:us
Real-time message bus for Claude Code sessions — one session asks another, the peer wakes on its own and answers, and the reply lands back in the asking pane. Bash + tmux.
The coordination layer for autonomous AI agents
Native Windows terminal multiplexer with cross-pane AI agent communication — no WSL2 required.
A Slack-style hub for humans and AI agents — real-time channels, external agent bots, file-aware chat, and channel memory. Open-source & self-hostable & MacOS & IOS.
DeepThink-AIOS is a fully local, multi-agent System with dual-sandbox code/logic verification, ChromaDB RAG memory, and dynamic VRAM multiplexing .
Local control plane for coordinating multi-agent work — tasks, claims, evidence, verification gates, and handoff briefs across sessions/agents.
Where local agent work becomes visible, durable, and malleable.
MALLO (MultiAgent LLM Orchestrator) es un sistema avanzado de orquestación de múltiples agentes de Modelos de Lenguaje de Gran Escala (LLMs). Diseñado para manejar consultas complejas, MALLO utiliza una variedad de agentes, desde modelos locales hasta asistentes especializados y APIs de terceros.
A local-first orchestration framework for OpenCode that coordinates planning, implementation, testing, review, retry, and escalation agents across software-development tasks.
Marblo — the live orchestrator for AI-native teams
Mission control for your agent swarm — know which agent is working, which is stuck, and which is done. Works with 14 coding agents, runs entirely on your machine.
Dispatch and review AI coding agents, hand them disposable VMs to test in, capture and summarize meetings and keep tickets, pipelines and your calendar on one deck. Available on macOS, Windows, Linux, browsers and phone.
DeepSeek Harness(dsh)动态 Cordis 插件:模型高可用回退 + 五种模式子智能体编排(fanout / pipeline / supervisor / map-reduce / router)
A gateway + skill letting Claude Code agents collaborate across machines
Claude Code instances on different machines can't talk to each other — SwarmCode gives them a Redis-backed channel for messages, shared artifacts, workspace status, and real-time push, plus a web dashboard.
Turn your devices into a fleet — dsh plugin for decentralized multi-device collaboration: mDNS discovery, key pairing, SSH direct exec. Tools auto-register for dsh agents.
Claude Code sessions and Codex CLI agents message one another by name, on one Mac or across your Macs over SSH. Native wake, receipts for every delivery.
Eureka — task-conditioned Meta-Agent orchestration for scientific discovery, developed by ManXis.
Build autonomous AI workflows with multi-agent orchestration, intelligent routing, hybrid RAG, knowledge graphs, MCP tools, persistent memory, and human-in-the-loop execution.
Where AI agents and people work together: shared channels and threads, tasks with one holder each, review by someone other than the author, and results shipped to GitHub, Slack and Linear, with anything else wired in by webhook. Hosted at maidan.world or self-hosted.
Cost-aware multi-agent orchestration for OpenCode CLI: route work to the right model, keep context small, and resume multi-session work from a Git-friendly plan board.
Local-first orchestration runtime for multiple autonomous coding agents (Claude Code, Codex CLI, Antigravity CLI, and more). A dynamic director plans, a deterministic kernel enforces.
Autonet: the daemon a household runs, the tool library, the services market and the contracts. The automation module of the On-Chain Jurisdiction.
A platform-agnostic, file-based multi-agent engine. Install 20 specialized AI personas directly into your repository. They plan, design, build, review, test, and ship — with structured memory that persists across sessions and two architectural moats no other framework combines.
Autonomous hierarchical multi-agent team (CEO -> Manager -> Workers) for Model Context Protocol (MCP). Mix Claude Code, Antigravity, Codex & Direct API with persistent SQLite queues and a real-time web UI.
LangGraph-based Supervisor-Worker multi-agent framework with a modern web GUI.
Communication between two Claude Code agents on different computers: they hand work over through Markdown files in a synced folder. Protocol, hooks and tests. ES + EN
Code and data for the Findings of ACL 2025 paper "Voting or Consensus? Decision-Making in Multi-Agent Debate". It compares seven decision protocols in multi-agent LLM debate: voting does 13.2% better on reasoning tasks and consensus 2.8% better on knowledge tasks.
Free subdomains under makes.fyi / agentdomains.co: landing page, docs and Cloudflare edge Workers for AgentDomains.
Experimental operations console for agent fleets. Derived from Paperclip.
COORDIVELA is the Development Coordination Control Plane for humans and AI agents: work state, leases, dependencies, schedules, conflicts, and interventions. COORDIVELA Open is its Apache-2.0 public surface: contracts, SDKs, examples, and approved core binaries.
Minimal, lightweight multi-agent skill + CLI for pi. No extension, no new tools. Run agents in parallel or in the background, then wait for, steer, stop or resume them. Works from any harness or agent that has a shell.
三省六部制多智能体审核流程,移植为 DeepSeek Harness 插件:12 状态状态机、门下省封驳、高风险确认、越权矩阵、委派防死锁、完整审计、军机处看板。
Oracle Generative AI Agentic Harness. Multi-agent reasoning orchestrator SDK with an admission gate in front of every consequential action, first-class on OCI Generative AI and Oracle Database 26ai.
The TeamAI application allows users to create a team of AI powered assistants with individual capabilities, personas. The AI assistants will solve the task requested by the user as a team effort, each bot contributing with its respective capabilities. Supported providers are Ollama and OpenAI.
The Cockpit for The Sovereign AI Software Factory