Loops HouseAEGIS — Autonomous Financial Governance & Security Infrastructure

NEAR HACKATHON
Team
1 member- TEOwner
Testimony
Overview
AI agents are increasingly capable of making financial decisions, but giving an agent access to money shouldn't mean giving it unrestricted control over how that money is spent.
We built AEGIS, an autonomous financial governance and security infrastructure that puts enforceable financial policies between AI agents and the transactions they initiate.
AEGIS allows individuals and businesses to register agents, assign spending allowances, establish shared budgets, restrict payment recipients and require human approval for sensitive transactions. It also provides treasury controls, duplicate-payment protection, an attack simulation environment and audit logs that explain why transactions were approved or blocked.
We integrated NEAR Protocol to move these controls beyond a conventional dashboard. Our NEAR testnet treasury demonstrates policy-controlled payments and rejected requests, while Chain Signatures enables policy-checked MPC signing for cross-chain transactions. NEAR Intents provides live cross-chain quotes, and NEAR AI Cloud supports payment-purpose reviews where available.
The prototype distinguishes between transaction authorization, signing and settlement, ensuring that an AI agent's request is never treated as permission to spend.
Our goal is to make autonomous financial systems safer and more accountable without sacrificing the capabilities that make AI agents useful