When ON, this page auto-opens a paper position every time the brain emits a finding meeting:
- Severity β₯ medium (2-star or higher)
- Model probability β₯ your threshold (default 70%)
- No existing position on that symbol
When the brain rates the outcome (30 min after firing), the position auto-closes. Wins = +risk Γ win-R, losses = βrisk Γ loss-R. Pure model performance β no human discretion in the loop. This is the cleanest test of model edge.