Every tool with its element, role, Test Status, key out-of-sample metrics, pass-bar reasons, and caveats — generated directly from runs/*.json. Click a row to expand the full audit detail. Statuses are the final Notion/HANDOFF labels, which in a few cases differ from the run files' proposed status (see the Findings page for the overrides).
| Element | Tool | Status | Role tested | OOS Sharpe (median) | OOS IC (median) | Max DD (median) | p-value |
|---|
Metrics are medians across the full parameter grid × walk-forward folds (see Methodology). p-values are bootstrap, uncorrected. "—" means the tool's role doesn't produce that metric (e.g. regime filters report accuracy + gated/ungated Sharpe instead).