Immune is a recursive testing and repair system for AI shopping agents. It tackles a subtle problem: a fix that stops an attack can also break legitimate purchases. A security agent proposes constraints against pricing abuse, while a customer advocate challenges policies that harm valid customers. When a fix fails, Immune spawns a child investigation and repair task. The revision must cite evidence from both agents and satisfy their constraints before undergoing further tests. Our live sandbox demonstration recovered from a deliberately seeded discount-policy failure, passing 8/8 private attack checks and 2/2 legitimate purchases in approximately 25 seconds. The engine supports two shopping-agent implementations we built. Crusoe powers inference, NVIDIA Nemotron provides independent review, You.com supplies research, and OpenShell executes sandbox checkout operations. An interactive canvas exposes task lineage, policy changes, advocate disagreements, and checkout receipts.
More from The Recursive Agents Hackathon
Etch turns one sentence into a real circuit board. You say “build a GPS tracker like an AirTag” and walk away. A team of agents picks real parts, writes the…
SPARKLETEAM is a self-improving agent team that teaches itself to load a dishwasher. It learned one law of physics and one superstition, and we measured which…
AGI Factory is a robot maintenance team that rewrites itself and keeps a change only when unseen factory months prove it better and safer. It runs a real…
Lets go : An agent that lives in your chat. Ask it to plan, and it turns everyone’s messages into rules. Our second agent, Mato, learns from LetsGo's…

Trendify is an AI-powered trend discovery platform that finds emerging trends before they go viral. It deploys thousands of autonomous agents across social…

Antaeus is an autonomous research lab for recursive agent improvement. Given a capability goal, Antaeus generates executable synthetic environments that…
