Strategy LabWhat it is — Chart-native Strategy Lab (v1 + DSL v1.5 + variant rail v2 + honesty kit v2.5): catalog / Condition DSL / multi-variant compare, plus buy-and-hold, seeded shuffle-null, and Go/No-Go on the same window and fill model.What it is not — Not TrendSpider Sidekick chrome. Not a parallel backtest engine. Not live trading / broker sync. Not walk-forward or holdout OOS (true OOS is later). Not NL/AI (v3). Not nested boolean trees / short DSL / non-allowlisted indicators. Not LuxAlgo Pine shipped as source. Not PineTS/AGPL / bots / broker CSV.Engine — POST /ta/backtest, /ta/strategy-lab/backtest, /ta/strategy-lab/compare, /ta/strategy-lab/honesty — all use app.scoring.technical.backtest._simulate. Honesty kit re-fetches OHLCV for the same ticker/days (does not reuse the prior Lab bar cache).Fill model — Signals use data available at the close of bar t; fills execute at the open of bar t+1 (no look-ahead). Round-trip transaction cost + slippage bps applied. Optional stop/target as fractions of entry. Shared across strategy, B&H, and null.Lookback — Daily bars; days param clamped 120–10000. Chart lookback uses the run calendar span (end−start), never session-bar count. Yahoo/provider history limits apply — soft-fail when bars empty.Limits — Compare max 4 variants; DSL long/flat. Honesty kit is in-sample only (null = shuffle of strategy signals, default seed 42). Go requires beat B&H + beat null + ≥5 trades + max DD leash. Not walk-forward.Lux / Pine — LuxAlgo Universal Signal Backtester ideas previously ported to lux_signal_backtester.py (CC BY-NC-SA — ideas only; their Pine is not shipped). Lab runs MRT catalog + Condition DSL + variant compare + honesty kit — not a Pine strategy() clone.Sources — MRT technical registry + Yahoo OHLCV via existing TA chart/backtest paths.