Skip to content
SWARM SQUAD

Open source · Multi-agent simulation

SWARM
SQUAD

A simulation framework for multi-agent systems. Three open source projects for building, breaking, and talking to autonomous swarms.

View on GitHubRead the Docs

Vehicle control · LLM decisions · Real-time dialogue

About

Testing a swarm in the real world is slow, expensive, and occasionally catastrophic. Swarm Squad is a controlled virtual environment for designing, running, and evaluating multi-agent behavior — formation control under jamming, agent-to-human dialogue, and everything in between.

The lineup

Three episodes, one framework

01The base

Swarm Squad

The foundational simulation framework — agents, environments, physics, and visualization for studying swarm behavior at scale.

“A simulation framework for multi-agent systems.”

PythonSimulation
02Episode I

Surviving the Jam

A hybrid control architecture pairing behavior-based formation control with LLM decision making — built for swarms that keep flying when communication gets jammed.

“Chaos erupts, AI swarms react… can they escape the gridlock?”

PythonLLM control
03Episode II

The Digital Dialogue

Real-time dialogue between agents and operators — WebSocket communication, a FastAPI backend, and a Next.js interface for talking to the swarm.

“AI agents speak, but can they truly understand?”

TypeScriptRealtime