openswarm-ai/openswarm

Your mission control center for a swarm of Ai agents.

What it solves

Open Swarm addresses the chaos of managing multiple AI agents simultaneously. Instead of juggling separate terminal windows and losing track of different workstreams, it provides a single, local interface to launch, monitor, and coordinate a "swarm" of agents working in parallel.

How it works

It operates as a local desktop application (Electron shell) with a FastAPI backend. The system uses a spatial dashboard—an infinite canvas where agents are represented as draggable cards. Each agent is isolated in its own Git worktree and branch to prevent conflicts. The orchestrator manages communication via WebSockets for real-time streaming and integrates with the claude-agent-sdk and MCP (Model Context Protocol) tool servers for capability expansion.

Who it’s for

Developers and power users who need to run multiple coding agents in parallel across different branches and require a centralized hub for tool approvals and conversation management.

Highlights

  • Spatial Dashboard: An infinite canvas for organizing agent cards, view cards, and embedded browsers.
  • Human-in-the-Loop: A unified approval workflow where all tool-use requests from all agents are centralized for a single click or keyboard shortcut approval.
  • Git Worktree Isolation: Each agent operates in its own dedicated git worktree and branch.
  • Conversation Control: Ability to edit prior messages to fork conversations and navigate between branches.
  • Extensible Tooling: Support for MCP tool servers (stdio, HTTP, SSE) and integration with the Anthropic skills marketplace.
  • Vibe Coding: Support for creating interactive HTML/JS/CSS artifacts rendered in iframes, which can be generated by the LLM.

Related

  • Project
    desplega-ai/agent-swarm

    An open-source operating system for AI work that coordinates a lead agent and specialized worker agents in isolated Docker containers to automate company-wide tasks.

  • Project
    agent-tower/core

    A command center for managing multiple AI agents, featuring a Kanban dashboard, automatic Git Worktree isolation for conflict-free coding, and multi-agent collaboration.

  • Project
    morapelker/hive

    An open-source AI agent orchestrator that allows developers to run multiple AI coding agents in parallel across isolated git worktrees.

  • Project
    VRSEN/agency-swarm

    A framework for building multi-agent AI applications that organizes agents into collaborative swarms based on real-world organizational structures.

  • Project
    swarmclawai/swarmclaw

    A self-hosted AI agent runtime and multi-agent framework for creating autonomous agent swarms with durable memory, scheduling, and multi-provider LLM support.