Helvetic Research

Most tools have clearly distinct purposes, but a few pairs are similar (e.g., get_drawdowns vs drawdown_series, run_quant_research vs run_strategy_v3, translate_strategy_request vs translate_and_run_strategy) and could cause misselection. Descriptions generally clarify the differences, making this a minor issue.

Naming is predominantly snake_case with a mix of verb_noun (e.g., get_metrics, run_backtest) and noun_analysis (e.g., correlation_analysis, regime_analysis). There are minor deviations like what_if_rebalance and full_research_brief that break the pattern, but overall it is consistent and readable.

With 85 tools, the surface is extremely heavy. Many tools are granular variants (e.g., daily_pnl_series, drawdown_series, trade_pnl_series) that could be consolidated, and the count far exceeds even a generous threshold for a single-purpose server.

The tool set is remarkably comprehensive, covering data retrieval, multiple backtesting modes, performance/risk analytics, optimization, robustness checks, strategy translation, reporting, and dashboard management. No significant workflow gaps are apparent.