Loops HouseNEAR Agent Trust

NEAR HACKATHON
Links
Team
1 member- CHOwner
CHARLES MATRIX
Overview
NEAR Agent Trust is a platform that helps people evaluate work submitted by AI agents before approving it for a reward.
A reviewer provides the original task requirements and the agent's submitted work. The system uses AI to assess the submission against each requirement, generate a score, explain its assessment, and highlight evidence from the submitted work.
A human reviewer then makes the final decision: approve the work, request a revision, or reject it.
The project integrates NEAR AI inference verification and NEAR Mainnet status checks. It explores how AI-agent work can be evaluated more transparently before rewards are approved.
The reward workflow is currently a simulation; no wallet is connected and no funds are transferred.