Every module, in the open

No black box.

What's actually running under the chart, how a signal becomes an order, and the full backtest disclosure. Nothing softened to sound better than it is.

01 · Detection

Liquidity Sweep Detection

Separates a sweep-and-reclaim from a genuine break, so only the reclaim feeds forward.

02 · Confirmation

MSS & Displacement

A structure shift only counts once a real range-and-close candle confirms it, not a wick.

03 · Mapping

Fair Value Gap Mapping

Plots imbalances as they form and tracks live fill state, aging stale zones out automatically.

04 · Decision

Arbitration Engine

The only module allowed to set trade direction; nothing else can act on its own.

05 · Filter

Volatility Regime Filter

A VEI-based read that hard-blocks entries during unstable, expanding volatility. Opt-in.

06 · Analytics

Live Probability Dashboard

Scores live setups against your own history, by session, regime and weekday.

07 · Risk

Risk Circuit Breakers

Daily loss limit, minimum order-size floor, and live consecutive-loss tracking.

08 · Timing

Killzone Highlighting

London and New York windows shaded live on the chart, plus an overlap bucket. Not just a background filter you can't see.

ASIA LONDON NEW YORK OVERLAP
09 · Execution

MT4/MT5 Webhook Automation

Confirmed signals fire a TradingView alert with entry, stop and take-profit already computed. Route it through your own webhook bridge and it lands as a live order in MT4 or MT5.

10 · Grading

Setup Grading

Every setup is scored against several independent internal models and reduced to a single grade from A+ to D. You see the grade, not the machinery behind it.

A+ A B+ B C D

Illustrative: grade reflects each live setup's own score.

Architecture

Built to hold under a full read-through.

  • Confirmed bars only Nothing evaluates on an unclosed candle.
  • One side at a time Long and short can never both be open at once.
  • Read-only analytics They never feed back into entries.
  • No lookahead Non-repainting patterns throughout.
On-chart dashboard

Every setup, scored
before you take it.

PROBABILITY DASHBOARD
BucketSampleWin %Bias
London Killzone14246.5%Bullish
NY Killzone11843.2%Bullish
London ∩ NY Overlap5451.9%Bullish
Regime · Compressed20144.8%Favorable
Regime · Expansion8931.4%Blocked

Illustrative: populates from your own chart history.

Sequence

Raw price to
placed order.

01

Detect

Sweep, MSS and FVG modules scan every confirmed bar independently.

02

Confirm

Displacement, regime and session all have to agree before a raw signal is valid.

03

Arbitrate

One engine resolves conflicting signals and sets the final trade direction.

04

Execute

Position sizing and loss limits run before an entry, stop or target is ever sent.

Execution

Alerts that place the order.

Every confirmed signal fires a TradingView alert with entry, stop and take-profit already computed. Route it through a webhook bridge and it lands as a live order in MT4 or MT5, no watching the chart required.

  • Pre-computed Stop-loss and take-profit are set by the engine and sent with the alert, not added by hand after the fact.
  • Bridge-agnostic Works with any TradingView-to-MT4/MT5 webhook bridge (e.g. TradingViewLink, TradersPost).
  • Non-custodial Lumen never touches your funds; the bridge and your own broker execute the order.
ALERT → EXECUTION
TradingView Alert

Fires on confirmed entry with stop & target attached

Webhook Bridge

Your own bridge service relays the alert to your terminal

MT4 / MT5 Order

Your EA places the trade with SL/TP already attached

REQUIRES YOUR OWN BROKER + BRIDGE SETUP

Two real samples.
Every window included.
Full disclosure, every time.

Full Disclosure

The backtest, exactly as it happened.

METHODOLOGY · SAMPLE 1 / LAST 12 MONTHS
251Trades sampled
54.98%Win rate (138/251)
1.417Profit factor
20.81%Max drawdown

XAUUSD, 15-minute chart, trailing 365 days from the TradingView Strategy Tester. Net return over the window was +123.58%, on 2,696.19 USD of max drawdown (20.81%), a real number we're not hiding because it's larger than the win rate looks comfortable with. 251 trades is a meaningfully bigger sample than a few dozen, though still short of the 500 to 1,000+ we'd want before calling an edge validated.

This replaced an earlier, smaller-sample baseline shown on this page, not because the old numbers were wrong, but because a full year gives a much better sense of how the system behaves across different conditions than 65 trades on a narrow window did.

METHODOLOGY · SAMPLE 2 / LAST 30 DAYS
25Trades sampled
64.00%Win rate (16/25)
2.192Profit factor
7.46%Max drawdown

XAUUSD, 15-minute chart, trailing 30 days from the TradingView Strategy Tester. A different timeframe and window than the baseline above, shown separately rather than blended together to look more consistent than it is. Net return over the window was +22.91% on 745.60 USD of drawdown. 25 trades is a small sample on its own; treat this as a recent snapshot, not a validated edge.

LIVE · AUTOMATED DEMO, VERIFIED BY MYFXBOOK

Everything above is a backtest. This is the same engine running live through the webhook bridge into MT5, on a demo account tracked independently by Myfxbook rather than self-reported, so the numbers aren't ours to edit.

Lumen live automated demo track record, verified by Myfxbook

Be clear-eyed about what this account can and can't tell you. A short forward-test window is exactly that: short. A handful of good weeks isn't statistically meaningful on its own, and treating it as proof of an edge is the same mistake as trusting a 20-trade backtest. Sample size and duration are what the 251-trade, 12-month backtest above is for. What a live demo account is for is different and narrower: confirming that the signal-to-order pipeline itself works exactly as described, with real (if simulated) broker execution and the losers left in, not just the winners. Treat the two as answering different questions, not the same one.

Demo account, not real capital. MetaTrader 5, 1:500 leverage, Vantage Markets. Verified badge and live figures update automatically from Myfxbook, not from us. View live on Myfxbook →

Comparison

Generic ICT scripts vs. Lumen.

CapabilityFree public ICT scriptsLumen
Instrument focusGeneral-purpose, any symbolPurpose-built and tuned for XAUUSD
Signal arbitrationFilters fire independentlySingle arbitration layer resolves conflicts
Displacement confirmationOften wick-basedRange-and-close required
FVG fill trackingStatic boxesLive fill-state tracking, auto-aged
Volatility regime awarenessRarely presentVEI-based hard gate, opt-in
Live probability contextNoSession / regime / weekday dashboard
Risk circuit breakersNoDaily loss limit, min-qty floor, loss-streak tracking
Setup gradingRare, if present at allA+ to D grade from multiple internal models
Broker automationRareWebhook alerts route straight to MT4/MT5
Published methodologyRareFull sample size & disclosure on this page
In Practice

What this changes about how you trade gold.

Not promises of profit. A different process.

Faster reads

Structure plots itself the moment it confirms.

Context first

Know your historical edge before you click.

Systematic cutoffs

The same risk rule applies, every single time.

Built for gold

Every threshold tuned for XAUUSD specifically.

Access

See the full pricing breakdown.

Monthly, quarterly and annual terms: every plan ships the complete engine, no feature gating.

View pricing

CONTINUE EXPLORING