HomeAI & Machine LearningContract Net Protocol: Multi-Agent Task Auction Simulator

Contract Net Protocol: Multi-Agent Task Auction Simulator (2D)

Watch a Contract Net Protocol auction unfold in 2D: a manager announces a task, worker agents bid based on load and distance, the lowest-cost bidder wins the contract, and a random-assignment baseline shows the real efficiency gain from auctioning.

AI & Machine Learning2DModerate60 FPS📱 Mobile-adapted⇄ 3D version
2d-ai-topic-14 ↗ Open standalone

The Contract Net Protocol is one of the oldest and most practical coordination mechanisms in multi-agent systems: instead of a central scheduler deciding who does what, a manager agent announces each task and worker agents bid on it based on their own estimated cost and current load, with the lowest bid winning the contract. This 2D canvas rendition draws the negotiation live on a flat plane — a manager node at the center broadcasts tasks to a ring of worker agents, visible bid beams show each worker's offer, the winning bid is highlighted and the task is awarded and executed, and the load-balancing effect of the load-penalty term becomes visible as busy workers price themselves out of new contracts. A running random-assignment baseline is scored alongside every real auction so the measured efficiency gain of market-based task allocation is visible, not just asserted. Tune the number of workers, task arrival rate, bidding noise and load penalty to see how the auction's efficiency and fairness change in real time.

⚙ Under the hood

Watch a Contract Net Protocol auction unfold in 3D: a manager agent announces a task, worker agents bid based on distance and current load, the lowest-cost bidder wins the contract, and throughput metrics update live.

multi-agent systemscontract net protocoltask allocationauction mechanismLLM agentscoordination

2D · HTML5 Canvas 2D · 60 FPS target · runs fully client-side, no install

What did you find?

Add reproduction steps (optional)