Free Claude skill built from 71 quant finance papers.
It does three things general AI knowledge does badly:
- Cites the page for every number. Buffett's realized Sharpe over
40 years was 0.79. That is the reference point most backtests
need and never get. - Runs the tests people skip. Deflated Sharpe, probability of
backtest overfitting, purged cross-validation, factor attribution
against live Ken French data. - Takes a side where the literature disagrees, and says what would
change its mind.
You can find it here.
