quantskills

PatternCatcher MCP skill for similar K-line stock and futures search

Detect backtest overfitting & multiple testing: Deflated Sharpe Ratio, PBO (CSCV), purged/embargoed CV, Harvey-Liu haircut. Research/education only, not investment advice.

PandaAI 因子大赛上手与在线挖掘技能：环境体检、登录、字段算子速查、可续跑批量回测与成本折算复盘 · Onboarding and online factor mining for PandaAI

SSQuant Agent workflow for futures strategies, data_server diagnostics, CTP/SIMNOW runtime checks, and Chinese backtest reporting.

Barra-style multi-factor risk model & risk attribution: cross-sectional WLS factor returns, Ledoit-Wolf shrinkage covariance, factor vs specific risk decomposition. Offline via a panda_data adapter. Research/education only, not investment advice.

Convex portfolio optimizer: mean-variance / min-variance / max-Sharpe / risk-parity / max-diversification with weight caps, sector & exposure neutrality and turnover limits. Research/education only, not investment advice.

Audit split and cash-dividend consistency across raw and adjusted equity prices before research or backtesting.

Stress portfolio liquidation capacity, pro-rata redemption shortfalls, and spread plus square-root impact costs.

Audit point-in-time universe membership, security lifecycles, stable identities, and missing delisting returns before backtesting.

Run reproducible index addition, deletion, and weight-change event studies around announcement or effective-date anchors.

Audit normalized intraday OHLCV data for timestamp, gap, price, volume, and trading-date defects before research or backtesting.

Audit continuous futures contract selection, roll events, same-day price gaps, and adjustment ledgers before research or backtesting.