個人 GitHub 帳號下公開儲存庫的擁有者可以登入並驗證所有權,自行管理介紹,無需管理員核准。
登入並產生驗證檔案,在 30 分鐘內將其提交至此儲存庫的預設分支。其他分支或分叉儲存庫中的檔案無效。然後在下方檢查檔案,即可無需管理員核准直接管理專案。請保留檔案以維持管理權限。
如需登入,將前往登入頁面。登入後請回到此處,再次點擊所選的驗證按鈕。
在專案資料夾中執行 /nomorevibe verify,即可驗證所有權並自行管理介紹。
一分鐘完成驗證正在載入完整 README。 以下為已儲存的摘錄。
Clean, long-horizon-only rebuild of the NiftyQuant long-horizon strategy: sma200_slope_63 top-15 cross-sectional trend-momentum on a PIT-clean, large+mid, solvent Nifty-500 universe.檢視完整 README ↗
Clean, long-horizon-only rebuild of the NiftyQuant long-horizon strategy:
sma200_slope_63 top-15 cross-sectional trend-momentum on a PIT-clean, large+mid,
solvent Nifty-500 universe.
nq/ package; transplant only the data corrections, empirical history, and golden
master that can't be regenerated from a spec.skills/ — pre-registration, n_trials/DSR, the
promotion bar, the baseline_v0 anchor. Binding.baseline_v1 — the pinned baseline of record (ADR-0006, dataset-pin-20260701).
sma200_slope_63 cross-sectional momentum, 63-day hold. Gross 15.46% CAGR /
Sharpe 0.667 / MaxDD −46% / ~12% after-tax. Byte-reproducible from the pinned
OHLCV. In-sample; no real capital has traded.weekly-swing 0094 — the weekly-swing book surfaced on the marketing site
("this week's scan"). NET of costs, corrected universe 2017–2026: 8.1× (₹10L → ₹81L,
+711%) / Sharpe 1.13 / CAGR 24.7% / MaxDD −42% / 59% win over 255 trades.
Reproduce: python scripts/run_bhanushali_weekly_rank.py. DSR 0.894 → below the
deflated-Sharpe gate → NOT certified. It is a registered proposal for the 2026-10-01
review + the forward wall — paper-tracked, not proven, before 20% STCG, no real capital.frontend/ — the marketing landing (src/pages/LandingV2.jsx) + the React
dashboard. The public landing serves aggregate, non-actionable stats only (no tickers,
no prices) via GET /api/landing-stats, with the "not certified · paper-tracked"
framing intact.dashboard/backend/ — FastAPI: signals, positions, trades, Kite (Zerodha) OAuth,
and the public landing_stats router (the frozen track-record / equity / weekly-scan
constants the landing renders).Numbers discipline: every figure on the landing is reproducible from the committed pipeline — never a chat transcript. Honesty framing (in-sample, deep drawdown, uncertified, no real capital) is load-bearing, not decoration.
nq/data OHLCV + PIT membership mask + features (sma200_slope_63, atr, ADV, D/E)
nq/engine panel + exits (shared) + portfolio.simulate (the backtest/live engine)
nq/strategy the long-horizon scan
nq/validation cpcv / bootstrap / dsr / metrics
nq/runner live scanner
research/ pre-registered findings (NNNN-*.md), overlay registry, exports
dashboard/ FastAPI backend (signals, Kite, landing-stats)
frontend/ React dashboard + marketing landing (LandingV2)