This page runs a deterministic simulation, not a model.
The policy below mirrors the bounded agent's tools, but the decisions are rule-based.
The published log carries "using_claude": false. A model-piloted run requires an API key and has not been published.