Team Idea
We’re building HUNT, an independent verification system for AI coding agents. AI agents can write code and pass tests, but that doesn’t necessarily prove they satisfied the original engineering requirement. HUNT independently challenges completed work, tests the actual requirement, inspects resulting system state, and produces evidence showing whether the task was truly completed correctly.
Our demo uses IBM Bob 2.0 to implement a real engineering requirement, then HUNT independently verifies Bob’s work rather than trusting the agent’s own tests or completion claim.
Team Members
Johnathon Slaughter
johnathon_slaughter565
Research and Dev Aviation
