🧪 Evaluation Harness Pipeline Simulator
Tasks flow from a generator through parallel agent workers to hidden-test oracles: tune arrival rate, worker count, agent skill and test strictness, and watch the scorecard update live.
AI & Machine Learning3DModerate60 FPS
⚙ Under the hood
Tasks flow from a generator through parallel agent workers to hidden-test oracles: tune arrival rate, worker count, agent skill and test strictness, and watch the scorecard tally pass/fail in real time.
Three.jsAIAgentsEvaluation
3D · Three.js / WebGL renderer · 60 FPS target · runs fully client-side, no install