# deep-strategy-optimizer Audit-and-fix convergence optimizer for trading strategies and their backtests: 19 passes over simulation integrity, statistical honesty, claim provenance, and economics — where 'no promotion' is the expected result. The strategy member of the family, built around one premise: most backtest edges are artifacts. Nineteen passes audit a strategy, its card, and the research behind it — simulation integrity (lookahead, cost path, accounting, data), statistical honesty (protocol grade, multiple-testing burden, evidence floor, degeneracy, attribution), claim provenance, economics, and tests. Medium+ fixes are applied and verified against the project's own suites, looping to convergence. The opt-in `--promote` mode runs a gated champion–challenger climb on held-out cross-validation blocks — where **no promotion is the expected outcome**, and a promotion has to earn its way past the gate rather than past the operator's optimism. **Use it for:** "run dso", "optimize this strategy", "is this backtest honest", pre-deployment audit of a strategy card and its evidence. **Not for:** the backtest *code*'s general quality ([code-deep-optimizer](/skills/code-deep-optimizer/)) · market research with no strategy artifact yet · portfolio advice.