Trust the process.
Inspect the evidence.

Reported fills, not broker statements.

Nyx’s System 01 journal reads the existing public Cloudflare closed-trade feed. TradingView alerts supply the records; this site does not verify account fills, alert completeness, or source immutability. The page refreshes every 45 seconds. Retrieval time is not a pipeline heartbeat.

The forward feed includes estimated commission but no slippage; the static backtest has different cost assumptions and is not out-of-sample validated. Neither is combined with Bhav’s research. Nyx’s supplied data lacks an admitted-session calendar and daily completeness authority, so standardized last-5/10/30-session economics are unavailable. We do not infer sessions, fill missing days with zero, or substitute last-N-trades windows.

One sealed research source.

Bhav’s archive is a separate, allowlisted projection of the same sealed research publication used by QuantNQ. It does not recalculate performance for sales copy or pull a live trading account feed.

The current public catalog contains 1,256 ledger views. The latest selected publication cutoff is Sep 25, 2026, but individual studies may end earlier. The cutoff printed on each record is authoritative for that record.

Sources and generated detail files are verified by byte count and SHA-256. These hashes prove snapshot consistency, not independent authenticity, an audit opinion, or research-to-runtime parity.

Unknown is not zero.

Calendar membership comes from admitted research sessions, not inferred weekdays. A missing complete-session outcome remains null. A resolved subtotal can exist on an otherwise unknown day without making that day complete.

Whole-window P&L is unavailable if any admitted session is unknown. The chart breaks and resets its cumulative path at every unknown session. Known no-trade or flat sessions remain in the record. Session shares use all admitted sessions, including unknowns.

No handpicked recent streaks.

Every Bhav research detail record exposes full history and the fixed last 5, 10, and 30 admitted sessions ending at that record’s cutoff, including exact dates and retained daily rows. These windows are descriptive only: they do not select a strategy, promote a candidate, or establish untouched forward evidence.

Dollar economics use each study’s published accounting. Costs, contract sizing, exposure, calendars, and sample sizes vary. Public aggregates deliberately omit private parameterization, so they are not sufficient to recreate a strategy or independently model brokerage returns. They are not capital-normalized returns, a promise of achievable fills, or proof that two strategies share the same risk.

Trade profit factor and daily profit factor are different measures. Missing support stays null; a supported positive, loss-free profit factor is shown as infinity. Win rate uses wins divided by wins plus losses, excluding zero-P&L outcomes. Best-day contribution can exceed 100% or be negative.

Maximum known-segment drawdown resets at unknown sessions and is observed at the published daily-close granularity. It is not intraday drawdown, worst-case risk, or a funding requirement. Where event economics are not supplied, event metrics remain unavailable rather than being inferred from daily totals.

A component is not a contribution.

Each portfolio is a distinct accounting authority. Standalone results cannot be summed into an overlap-accounted parent. A linked component may have a separate cost variant from its featured standalone view; exact variants retain separate public identities.

Options-flow alternatives are not an options-inclusive joint portfolio. Unevaluated combinations are listed as unavailable. A comparison page neither sums nor optimizes its selected records. The expanded Bhav collection includes seven current complete-positive histories and six separately labeled incomplete research records. Current continuity identities retain an original prefix plus a disclosed source-policy suffix; they are not homogeneous retrospective backtests. Older IDs preserve the original evidence. It is not an out-of-sample validation or an automatic promotion based on recent windows. Session Reversion retains its all-history-selection and failed replacement-gate caveats. The full archive remains available.

A scenario, not a firm promise.

The performance desk reuses the repository’s daily-close account-analysis engine. Its fixed-floor and uncapped EOD-trailing scenarios keep each strategy’s original published size and trading costs. Evaluation and evaluation-to-first-payout use one attempt and at most 120 admitted sessions total, with explicit hypothetical fees, consistency and withdrawal rules. They are not official firm products, intraday risk checks, or R2’s account model.

Every admitted session starts a replay. All starts remain in the denominator, including shortened tails. Observed successes, breaches and unfinished outcomes stay separate. Medians condition on reaching the event; overlapping starts are not independent trials or estimates of future probability. Any unknown economics or missing accepted-activity proof blocks the entire strategy’s scenario output.

The separate scenario artifact binds the exact catalog, detail, FULL/L5/L10/L30 windows, frozen scenario config and analysis-engine hashes. No contract scaling, risk optimization, bootstrap or promotion is performed. Updated sessions require a new verified public export and an atomic app/data deployment; web requests do not replay research.

Evidence, not the recipe.

The Bhav research projection receives neutral public identities, approved labels, daily aggregate outcomes, window metrics, and receipts. It does not receive execution rules, parameters, entry/exit timestamps or prices, stops and targets, raw trade ledgers, research source paths, broker credentials, Pine code, private reports, or provider ladders.

Aggregate economics still disclose historical behavior and could support inference. Publication requires owner review and appropriate redistribution rights; sanitization is not a guarantee that all strategy characteristics are unknowable.

Research does not place orders.

Packaged references, portfolio and monitoring specifications, installed execution plugins, and practice-authorized products are separate states. No Bhav research catalog item is practice-authorized or live-authorized. Nyx’s external TradingView subscription is a separate tool offering, not an AlgosPaid execution plugin. Historical results are not attributed to a related runtime plugin without exact parity.

Checkout, when enabled for a specifically approved research offer, buys only the disclosed research deliverable. It never activates the trading bot. Practice execution would require a separately reviewed immutable plugin, server entitlement, verified practice account, local enablement, reconciliation, and centralized risk approval. Funded and live execution remain prohibited.