{"service":"x402-insights","what":"Trading-research verdicts, regime reads, and labeled trade archives from a live multi-platform prop-trading operation. All performance rows carry n + evidence class (live_fill > forward_test > paper/sim). No raw market data.","network":"base","pay_to":"0x4e90cc3BCCA166E3A322D524A5c5BB28ec4C6Abe","endpoints":[{"path":"/premarket","price":"$0.01","description":"Daily premarket volatility & regime read for US index futures/0DTE: vol_score, regime label, expected range %, primary driver, key events, reasoning — plus NQ overnight enrichment (overnight range pts, gap pts + direction, 20-session range percentile, heuristic signals; point deltas only, refreshed every 20 min). ~60 sessions of history."},{"path":"/calendar","price":"$0.01","description":"Normalized red-folder economic calendar: date, time ET, event, impact, forecast, previous."},{"path":"/social-vol","price":"$0.01","description":"Truth Social post -> volatility-explosion classification stream (flag + confidence + keywords). Event-driven."},{"path":"/daily-brief","price":"$0.01","description":"Last-10-day P&L brief across three live trading platforms (NQ futures, multi-asset futures, options), per strategy, evidence-class labeled (live_fill / forward_test / paper / sim). Price-stripped."},{"path":"/verdict","price":"$0.05","description":"Strategy verdict lookup: 165 registered strategies with documented test outcomes — tick-replay PF, forward-test results, real-money fill stats (n, PF, strip-2-best-days, split-half), and the written verdict. The BT->live gap database: what survived real money and what didn't."},{"path":"/archive","price":"$0.05","description":"Bulk labeled trade-outcome archive: 10k+ rows across NQ futures, 9 multi-asset futures symbols, 44 options strategies. Price-stripped, outcome + evidence-class labeled. Ensemble training data."},{"path":"/archive/edge-now","price":"$0.03","description":"Conditional entry edge: strategy x regime x time-of-day bucket x realized-vol tercile -> historical PF/win-rate/trade-count for exactly the current intraday conditions. GROUP BY over the /archive live-fill database (evidence_class=live_fill only), joined against /premarket's regime + vol_score history. edge_vs_baseline is the delta vs the strategy's unconditional /verdict live_fill_pf; entry_urgency is PRIME/FAVORABLE/SUBOPTIMAL/NO_DATA (<5 historical analogs -> revert to unconditional). Dimensionless ratios and labels only, zero price levels. Per-entry companion to /decide/next-trade."},{"path":"/opening-range","price":"$0.01","description":"NQ opening-range read (09:30-09:45 ET): OR width in points + 20-session percentile, first-bar direction, open vs overnight midpoint delta, overnight high/low breach flags, heuristic signals. Companion to /premarket — the session-confirmation half of the buyer loop. Point deltas and labels only; refreshed every 20 min."},{"path":"/fitness","price":"$0.05","description":"Regime-conditioned strategy fitness: for each NQ strategy, REAL win-rate/PF computed on live-money fills only, broken out by which regime /premarket assigned that session. Thin samples are honestly labeled insufficient_data rather than scored. Companion to /verdict — today's regime cross-referenced against strategies' actual historical performance in that regime."},{"path":"/regime-update","price":"$0.02","description":"Intraday regime-drift check vs the morning /premarket call: realized vol over the last 15 min compared to the same time-of-day distribution over the last 20 sessions, flagging whether the morning regime call is still valid. Refreshed every 20 min during RTH; returns session_active:false outside 09:30-16:00 ET."},{"path":"/prop-proximity","price":"$0.02","description":"Real-time danger score for a funded/eval prop account: given account state (trailing high balance, current balance, today's P&L), computes breach distances against drawdown/loss limits. Verified numeric thresholds for LucidFlex/LucidDirect (firm=lucid_flex_50k etc, see ?firms for the list); any other firm requires you to supply trailing_dd_limit_usd/daily_loss_limit_usd explicitly — we do not guess prop-firm rules we haven't verified ourselves."},{"path":"/prop-eval","price":"$0.02","description":"Evaluation-phase progress tracker: given a firm and a list of daily P&L, computes profit-target completion and pace. Verified profit targets for LucidFlex (?firms for the list); other firms require profit_target_usd explicitly. No fabricated pass-probability or ETA — those need a real model we don't have yet."},{"path":"/prop-rules","price":"$0.05","description":"Prop-firm execution intelligence from months of live funded/eval trading (Apex, Lucid, NinjaTrader/Tradovate): trailing-drawdown mechanics that kill accounts, phantom-fill and silent-stop-rejection detection, payout math, holiday lockout windows. Every rule carries its evidence. Filter by ?firm= or ?category=."},{"path":"/prop-portfolio","price":"$0.05","description":"Multi-account prop portfolio aggregator: given a JSON array of accounts (firm, eval_phase, account_size_usd, start_date, daily_pnl), runs /prop-eval + /prop-proximity-style arithmetic across every account and adds the cross-account view a single-account call can't produce — consolidated drawdown exposure, simultaneous-loss-day correlation risk, and a per-account today's-risk recommendation. Verified thresholds for LucidFlex/LucidDirect firms (see /prop-rules ?firms); other firms accept explicit profit_target_usd/trailing_dd_limit_usd/consistency_rule_pct per account."},{"path":"/transform/pdf-to-markdown","price":"$0.01","description":"Convert a PDF to clean LLM-ready Markdown (layout-aware, tables, OCR fallback for scanned pages). POST JSON with either 'url' (http/s link to a PDF, max 15MB) or 'pdf_base64'. Returns markdown + page count. Stateless — documents are processed in memory and never stored."},{"path":"/premarket/globex","price":"$0.02","description":"Overnight Globex gap/regime-alignment classifier: gap direction + size in ATR multiples, overnight range percentile vs last 60 sessions, overnight trend direction, close-in-range position, and whether the gap ALIGNED/CONTRADICTED this morning's /premarket regime call (with continuation bias + historical gap-fill rate before 11 AM ET). ATR multiples and percentile ranks only, zero price levels. Companion to /premarket and /opening-range."},{"path":"/fitness/risk-adjusted","price":"$0.05","description":"Risk-adjusted return statistics per NQ strategy: Sharpe/Sortino/Calmar computed on daily-aggregated REAL live-money fills (evidence_class=live_fill only), trailing 90 calendar days, plus win_rate, max_drawdown_pct, vol_of_returns, avg_win_loss_ratio, and a regime-conditional Sharpe filtered to today's /premarket regime. Thin samples honestly labeled sufficient_sample:false. Dimensionless ratios and percentages only, zero price levels. Companion to /fitness — pairs regime-active strategies with how well they've actually paid per unit of risk."},{"path":"/decide/next-trade","price":"$0.03","description":"Intraday per-trade approval skill: composes /regime-update, /fitness, /calendar and (when balance/trade state is supplied) /prop-proximity-style checks into one bounded GO/NO-GO with approved_size and a reason string. valid_for_seconds:60 — call once before every order. No account IDs handled or stored; balance/trade-count state (if you want the prop-account checks included) is supplied by the caller per-call, same pattern as /prop-proximity. Sibling to /prop-portfolio for the intraday loop-position decision instead of the eval-tracking one."},{"path":"/prop-health","price":"$0.02","description":"Session-level risk score for a prop/funded account — no strategy, direction or contract count required (that's /decide/next-trade's job). Composes /regime-update (drift + intraday vol label), /fitness (share of registered strategies UNFIT in today's live regime), /calendar (high-impact event proximity), /social-vol (recent explosion-flagged posts) and, when balance state is supplied, /prop-proximity-style danger score into one 0-100 session_risk_score + label (normal/watch/elevated/critical) + recommended_action. Heuristic v0 weights, disclosed as such; components with no data available are dropped and remaining weights renormalized rather than guessed. No account IDs, no price levels."},{"path":"/calibration","price":"$0.10","description":"How good are our numbers, measured: premarket range forecast vs REALIZED range per session (n, MAE, bias, hit-rate — updated every snapshot, misses included), plus methodology constants: the 2-7x bar-backtest PF inflation multiplier, strategy survival rates, evidence hierarchy. Calibration data for anyone consuming trading forecasts or backtests."}],"free":["/","/health"],"honesty":"Of ~200 strategy variants run for 4 months on real prop-firm money, exactly ONE cleared every validation test. We publish the failures. Verify our numbers — they hold."}