DETTSEC ALGO PVT · A DETTSEC PRODUCT← BACK TO DETTSEC
A Trading System Thesis · Binance USD-M Futures

Everyone else draws the smart-money setups. This one trades them.

SmartRisk is an automated "Smart Money Concepts" (SMC) trading engine for crypto. It reads a price chart the way an institutional trader does — spotting order blocks, fair-value gaps, liquidity grabs and trend shifts — then scores its confidence, manages its own risk, and can place the trade for you. The popular tools on the market only draw boxes on your screen and leave you to trade by hand. This is the full breakdown of what it does, how it works, and where it fits in the market.

127
codified smart-money strategies, distilled from 126 trading-course videos
15+
institutional chart concepts detected automatically, in real time
<1ms
C++ execution core — physics-timed order firing
$25T
Binance Futures volume in 2025 — the market this serves
The one-sentence version

An institutional chart-reader, a physics engine, and a risk manager — fused into one auto-trader.

"Smart Money Concepts" is the most popular way retail traders try to trade like institutions: follow the footprints big players leave on the chart. The problem is it's slow, manual, and emotional. SmartRisk turns that hand-drawn art into a systematic engine that watches the whole market at once, waits for high-confidence setups, and executes with strict, disciplined risk.

🎯

It reads the chart

Automatically finds order blocks, fair-value gaps, liquidity sweeps (stop hunts), trend breaks and the "premium vs discount" zones — the exact concepts SMC traders draw by hand.

🚀

It feels the momentum

A physics-inspired "kinematic engine" treats price like a moving object — measuring its speed, acceleration and force — to catch the exact instant a move ignites, with a C++ core that fires in under a millisecond.

🛡️

It manages the risk

A "100×" risk engine keeps every trade to a small fixed risk while freeing up capital, places smart stops behind liquidity, trails to lock profit, and shuts down after a losing streak.

How to read this document. It's a complete, plain-English tour of the system: the live dashboard, how it reads a chart, the physics engine, the decision brain, the risk rules, the problems it solves, the market opportunity, and a business case. Every trading term is explained as we go — no prior knowledge needed. Where a part of the system is still a blueprint rather than shipping code, we say so clearly.
The Dashboard · faithful recreation

The command centre

Below is a faithful recreation of the real GPU-accelerated dashboard, rebuilt from the application's own interface code — the same Binance-dark theme, the same gold/green/red colour language, the same panels. It scans the entire futures market live, and the whole risk engine is tuned from the bar at the top.

Binance Top Movers — GPU Accelerated · 1400×900
📡 BINANCE FUTURES MARKETS — REAL TIME DASHBOARD
▶️ START AUTO-TRADER
📈 MARKET SCANNERS
💼 ACTIVE PORTFOLIO
⚙️ RISK ENGINE SETTINGS Risk Per Trade (%): 2.0 Take Profit (%): 100 Stop Loss (%): 50
⚡ TRAILING STOP SYSTEM Trail 1 Trigger: 100 Trail 1 Stop: 0 Trail 2 Trigger: 200 Trail 2 Stop: 100

🔥 TOP GAINERS (24H)

SymbolPrice15m
MOODENG0.2841+7.2%
PNUT1.284+5.1%
WIF2.014+3.8%
FARTCOIN1.512+2.9%
POPCAT0.512+2.1%

🩸 TOP LOSERS (24H)

SymbolPrice15m
ORDI12.44-6.4%
JASMY0.0192-4.7%
SEI0.284-3.6%
TIA3.812-2.8%
AR6.14-2.2%

👑 TOP VOLUME (24H)

SymbolPrice15m
BTCUSDT98,240+0.4%
ETHUSDT3,412+0.8%
SOLUSDT214.8-0.6%
XRPUSDT2.284+1.2%
DOGEUSDT0.384+0.9%

⚡ TOP MOVERS (15m)

SymbolVolume15m
MOODENG$84M+4.1%
HYPE$61M+3.3%
ORDI$52M-3.0%
PNUT$47M+2.6%
SEI$39M-2.4%
↑ Real palette: background #0b0e11, cards #181a20, Binance gold #FCD535, bullish green #0ECB81, bearish red #F6465D. Illustrative live data.

Two screens, one engine

The Market Scanners tab (shown above) watches every USDT futures pair at once and ranks them into four live lists — biggest gainers, biggest losers, highest volume, and the fastest 15-minute movers — updating a few times a second. The Active Portfolio tab (right) is the account view: total balance, live profit/loss, and every open position with its side, entry, mark price and estimated return, all colour-coded green for profit and red for loss.

Above both sits the Risk Engine bar: risk per trade, take-profit, stop-loss and a two-tier trailing system — the whole money-management policy, editable in one place. One button hands the wheel to the auto-trader.

Active Portfolio
Total Balance
$4,820
Unrealized PnL
+$312
Active Trades
3
SymbolSideEntryMarkPnLROI
WIFUSDT🟢 LONG1.9422.014+148+74%
SEIUSDT🔴 SHORT0.2960.284+96+40%
TIAUSDT🟢 LONG3.883.812-32-18%

A second tool: the Funding-Rate Scanner

A separate window hunts for "funding" extremes — small recurring payments between long and short traders. When funding goes deeply negative, shorts are paying longs, which can flag a squeeze setup. It lists every coin sorted by the most negative funding (dark-red = extreme) and charts the history as a cyan line.

Binance Funding Rate Scanner (Extreme Minus)
SymbolFunding %Who Pays WhomStatus
MOODENG-0.182%Shorts → LongsTRADING
PNUT-0.094%Shorts → LongsTRADING
WIF-0.061%Shorts → LongsTRADING
BTC+0.010%Longs → ShortsTRADING
Funding Rate History: MOODENGUSDT
Smart Money Concepts · how it reads a chart

Seeing the footprints of the big players

"Smart Money Concepts" (SMC) is the idea that large institutions leave repeatable footprints on the chart — and if you can read them, you can trade alongside the big money instead of getting run over by it. SmartRisk detects all of these automatically. Here's a live chart with the engine's markings, followed by a plain-English guide to each one.

Up candle Down candle Bullish Order Block Fair Value Gap Liquidity / sweep

An illustrative recreation of how the engine annotates a chart: it marks the liquidity pool, the stop-hunt sweep, the change-of-character, the fair-value gap and order block in the discount zone, and the entry / stop / target — exactly the concepts below.

🧱

Order Blocks

The last candle before a big institutional move — the "footprint" of where smart money loaded up. Price often returns to it. The engine only trusts a block if the move that followed was unusually powerful (a filter called displacement), and treats it as live until price closes through it.

🌫️

Fair Value Gaps

A gap left by a violent three-candle move where price skipped a range. Markets tend to come back and "fill" these gaps, which makes them high-probability entry zones. The engine only counts a wick into the gap as an entry, not an invalidation.

💧

Liquidity & Sweeps

Clusters of stop-loss orders sit above equal highs and below equal lows. Big players spike price just past them to grab those stops — a stop hunt — then reverse. A confirmed sweep is the engine's highest-probability reversal trigger.

🔀

Break of Structure & Change of Character

BOS = price keeps going and breaks a prior peak (trend continues). CHoCH = the first break in the opposite direction — the earliest hint the trend is flipping. This is the #1 early-reversal signal in SMC trading.

⚖️

Premium vs Discount

It splits the recent range in half: only buy in the cheap "discount" bottom half, only sell in the expensive "premium" top half — and marks the 61.8–78.6% "sweet spot" (the Optimal Trade Entry) for the best risk/reward.

🎣

Inducement

A small, obvious "bait" level placed before the real zone, designed to lure breakout traders in so smart money can run their stops. Knowing where the trap is tells the engine where the fake-out will happen before it happens.

Two things almost no retail tool does

Rare / advanced

It measures the market's "heartbeat"

Instead of a fixed setting like "a swing is 5 candles," it uses radio-style signal processing (a Hilbert-transform cycle detector) to measure the market's current rhythm — fast chop or slow swings — and auto-adjusts. This "zero-parameter" design means the same engine works on Bitcoin and on a fraction-of-a-cent memecoin without re-tuning.

Pro-level ICT concept

It cross-checks against Bitcoin

It compares each coin against Bitcoin and Ethereum (SMT divergence): if an altcoin makes a new low but Bitcoin doesn't, that hidden strength is a bullish tell. It also widens your stop automatically when candles are wicky, so you don't get "wicked out" of a good trade.

Under the hood these run on Numba just-in-time compilation — the heavy pattern maths is compiled to near-C speed so the engine can scan hundreds of coins continuously.

The Kinematic Engine · physics for price

Treating price like a moving object

This is the system's most unusual idea. Most tools ask "is price going up?" The kinematic engine asks the questions a physicist would ask about a thrown ball: how fast is it moving (velocity), is it speeding up or braking (acceleration), is that changing (jerk), and how much force is behind it (money = mass). Just as physics can predict where a ball lands, these motion measurements predict where a move is about to ignite — and because the maths is simple arithmetic, it runs on every coin, every second.

Price (top) with its velocity and acceleration underneath. When force overwhelms the market's normal "noise," the engine fires — that ignition point is what it's built to catch.

Built as a four-part relay for speed

The engine is split into four independent programs (called "nodes") that pass work down a chain through shared memory, so the slow, deep analysis never holds up the instant execution.

① Macro Brain reads the chart every 15 min → leaves an SMC "cheat-sheet" ② Vanguard Ingester streams live prices for every coin into a 1-hour RAM tape ③ Kinematic Matrix the physics — speed, force, friction — decides to fire ④ Striker (C++) a hyper-fast native program fires the order in <1ms Shared memory relay — slow analysis never blocks fast execution

The three "ignition" patterns it hunts

💥

Panic Liquidation

Near a support level, a violent flush of selling is instantly absorbed and price snaps back up — the "dip then rip." The engine catches the exact reversal.

🐢

Stealth Grind

Quiet, low-force accumulation that steadily builds — speed, acceleration and jerk all creeping positive — right before a breakout. The calm before the launch.

🪤

Inducement Trap

Price is lured just past support to grab stops, then brutally reverses. The engine reads the snap-back as a high-conviction entry.

The "anti-trap" filters. To avoid firing on retail noise, the physics core uses statistics that adapt to each coin — including Otsu's method (borrowed from image processing) to separate institutional force from ordinary chop, and market "entropy" to know when a coin is coiled versus random. If huge volume is being absorbed with no price movement, it flags an absorption trap and stands aside.
⚙️

Why a C++ core?

The final "striker" is written in native C++ running a tight watch-loop on shared memory. The instant the physics engine writes a buy or sell, it signs and fires the order to Binance with essentially zero delay. In crypto, the first mover gets the fill — this is the low-latency edge that a normal Python bot can't match.

🔔

Alerts on every strike

Each signal also fires an instant Telegram alert, so you see what the engine saw — even when you're away from the screen. The striker ships in a safe simulation ("mock") mode by default, so it proves itself before it ever touches real money.

The Brain · from courses to code

How 126 trading videos became a trading engine

The knowledge that powers SmartRisk was distilled from a library of 126 professional Smart-Money trading-course videos. Each concept was turned into a structured, repeatable strategy — and those feed a decision engine that only pulls the trigger when enough independent signals agree.

📚 126 SMC videos the knowledge base 📋 127 strategies codified, step-by-step 🎯 Confluence score only high-conviction fires ⚡ Entry + execution precise price, risk, order

The Confluence Engine — the gatekeeper

This is the part that actually decides whether to trade. Its rule is simple: never trade unless enough independent signals agree, and never fight the big-picture trend. It checks the 4-hour direction, refuses to buy right under a major ceiling (or sell right above a floor), then adds up a weighted score from the 1-hour trend, the premium/discount zone and the confluence of signals. Only if the score clears a bar — a bar that automatically lowers when a genuine institutional surge shows up — does the trade pass.

🚦

Trend gate

No longs when the 4-hour trend is down; no shorts when it's up. Trade with the big money, not against it.

📐

Space-to-move gate

It won't buy straight into a nearby resistance wall — there has to be room for the trade to actually run to target.

🎯

Precise entries only

The Entry Engine enters at a real smart-money level — a gap, an order block, or just past a swept level — and refuses to "catch a falling knife" with no structure.

It remembers what it's watching. An "entry memory" keeps track of setups it's waiting on, trades it has already taken, and setups it recently rejected — so it never fires the same trade twice and never forgets a setup that's still developing.

The bigger vision: a 300-feature "Cognitive Decision Engine"

Beyond the shipping engine, the project lays out an ambitious roadmap: turn the "art" of expert chart-reading into 300 individual measurable numbers an AI can learn from — where every market phenomenon becomes its own tiny, swappable calculator, all feeding a single machine-learning decision.

Phase 1

Liquidity & Traps

Double-sweep traps, inducement engineering, session-handoff traps, volume forensics, "coin personality" vs Bitcoin.

Phase 2

Zone Selection

Optimal-entry traps, deep-discount penetration, breaker-block flips, pattern-deception traps, killzone timing.

Phase 3

Precision

Exact stop/entry/target maths — wick-buffer formulas, tick-sniping avoidance, split-entry ladders, reward-to-risk gating.

Phase 4

Risk Gating

Black-swan halts, correlation gates, chop-zone shutdown, funding-rate filters, dynamic position sizing.

Phase 5

Account Psychology

Drawdown circuit-breakers, euphoria brakes, revenge-trade denial, house-money scaling, expectancy feedback.

The design

Infinitely upgradable

Each feature is one isolated file. Invent a new market trap in 2027? Drop in one file and the engine loads it automatically — no rewrite.

Honest status: what ships vs what's the roadmap. The working, shipping engine today is the disciplined stack above — the SMC chart-reader, the kinematic engine, the Confluence gatekeeper, the Entry engine, the live Binance executor, the risk engine, and the full library of 127 written strategy blueprints distilled from the video knowledge base. The 300-feature Cognitive Decision Engine, the machine-learning layer, and the 127-strategy parallel "hive" are a well-documented architecture that is still mostly a blueprint — the plumbing (the plugin loader, the base templates, a first prototype strategy and a first feature) is built and proves the pattern works, but the full 300 features and the multi-strategy swarm are a roadmap, not a running system yet. We flag this so the vision and the shipping product are never confused.
The Risk Engine · "100×"

Where "100×" really comes from

The name sounds like reckless leverage. It's the opposite. The core idea is to separate risk from leverage completely. Your risk is set by where your stop-loss is — not by the leverage. High leverage is used only to lock up less collateral, freeing capital for other trades, while every single trade still risks the same small slice of the account.

The 2% rule, in one worked example

Balance $1,000 · risk 2% · entry $100 · stop $98

Risk $20 2% of $1,000 Size 10 coins $20 ÷ $2 stop Position $1,000 10 × $100 notional Margin locked: only $20 at 50× leverage — control $1,000, risk $20

You command a $1,000 position, risk exactly $20, and tie up only $20 of collateral. Same risk, far more capital efficiency — and, the design argues, it would take hundreds of losses in a row to drain the account.

Protecting profit automatically

🪜

Two-tier trailing stop

When a trade reaches +100% return, the stop moves to break-even — the trade is now risk-free. At +200%, the stop jumps up to lock in +100%. Winners get room to run while profit is protected. Crucially, positions keep defending themselves even when the auto-trader is paused.

🛡️

Smart-money stop placement

Stops don't sit at round, obvious levels. An "inducement shield" hides the entry behind the bait level so a stop-hunt can't clip you first, and the stop "breathes" with volatility — tighter in calm markets, wider in violent ones — so you're not wicked out of a good trade.

Hard safety rails

📉

2% per trade

A hard cap on how much any single trade can lose.

🔴

3-loss shutdown

Three losses in a row and it stops for 24 hours — no revenge trading.

🚧

Daily drawdown limit

Lose more than 3% in a day and trading halts until tomorrow.

⚖️

Minimum 1:2 reward

A trade is rejected unless the potential reward is at least twice the risk.

Why prop firms will like this. Funded-trader "prop firm" accounts live or die on strict drawdown limits and rule-consistency — exactly what an automated, fixed-risk, circuit-breakered engine delivers. Discipline that humans struggle to hold, the machine holds every time.
Problems It Solves

Smart Money Concepts works — but trading it by hand doesn't

SMC is one of the most popular trading methods in the world, yet most people who try it lose. Not because the concepts are wrong, but because executing them manually is slow, emotional, and impossible to do consistently across hundreds of coins. Every feature in SmartRisk removes one of those failure points.

The problem

The popular SMC tools (like LuxAlgo) only draw order blocks and gaps on your chart. You still have to watch, decide, and place every trade yourself — by hand.

The solution

It closes the loop. SmartRisk detects the setup, scores it, sizes it, and can place the trade on Binance automatically. From "here's a box" to "the trade is on."

The problem

SMC trading is emotional. People miss setups, hesitate at the entry, move their stops, and revenge-trade after a loss. Discipline is the hardest part.

The solution

A machine has no emotions. It follows the rules every time, and hard circuit-breakers (3-loss shutdown, daily drawdown limit) enforce the discipline humans can't.

The problem

Not every order block or liquidity level is real — many are traps that lure you in and stop you out. Beginners can't tell the difference.

The solution

It filters for quality. Zones are only trusted if the move that made them was powerful (displacement), it maps the "inducement" traps, and it demands multiple signals agree (confluence) before acting.

The problem

No human can watch hundreds of coins across multiple timeframes, 24 hours a day. The best setups happen while you're asleep.

The solution

It never sleeps. It scans every USDT futures pair continuously, and fires an instant Telegram alert the moment a setup triggers — so nothing slips past.

The problem

Every coin behaves differently. Settings tuned for Bitcoin fall apart on a fast memecoin, so fixed-setting tools constantly need re-tuning.

The solution

Zero-parameter design. It measures each coin's own rhythm and volatility and adapts automatically — the same engine works from Bitcoin to a fraction-of-a-cent altcoin.

The problem

Leverage blows accounts. Most traders confuse "high leverage" with "high risk" and wipe out on a single bad trade.

The solution

Risk is decoupled from leverage. Every trade risks the same small fixed slice of the account regardless of leverage, with a minimum 1:2 reward gate on every entry.

The problem

Reacting a few seconds late means a worse fill — or missing the move entirely. Manual traders are always behind the ignition.

The solution

Physics-timed, C++-fast. The kinematic engine catches the exact instant a move ignites and a native execution core fires in under a millisecond.

The problem

Passing a funded-trader "prop firm" challenge requires flawless risk discipline over dozens of trades — most people fail on the rules, not the strategy.

The solution

Rule-perfect by design. Fixed risk, automatic drawdown limits and consistent execution are exactly what prop challenges reward — and what the engine does natively.

Market Analysis · July 2026

A huge audience, a tooling gap wide open

Smart Money Concepts is a mass movement, but the tools around it stop at "drawing boxes." Nobody combines detection, confluence scoring, machine learning and auto-execution across a whole strategy library. That gap is the opportunity.

2M+
YouTube subscribers of ICT — the originator of the SMC method
1.3M
followers of LuxAlgo, the leading SMC indicator maker
$85.7T
crypto derivatives traded in 2025 (~73% of all crypto volume)
2.1M
funded prop-firm traders — a prime buyer channel

Sources: ICT YouTube; TradingView/LuxAlgo profile; CoinMarketCap Academy 2025 derivatives report; Track360 2026 prop-trading report.

Everyone else draws. Nobody trades.

The whole field splits into two camps: indicators that detect and draw SMC setups then leave you to trade by hand, and generic bots that auto-trade but are completely blind to smart-money concepts. A handful of MetaTrader "SMC robots" sit in between — but they're single-strategy, fixed-rule, and built for forex, not crypto.

CapabilityLuxAlgo / TV indicatorsMT5 SMC robotsGrid / DCA botsSmartRisk
Detects SMC patterns (order blocks, FVG…)✓ draws thema few✓ full library
Auto-executes the SMC setups— alerts only1 strategyno SMC logic✓ Binance, C++
Confluence scoring across strategies✓ ~127
Machine learning on SMC features— backtester only— fixed rulesgeneric "AI"✓ 300-feature (roadmap)
Dynamic, confluence-based riskstaticgrid sizing✓ adaptive
Crypto-native (Binance USD-M)chartsforex/MT5

Based on official product pages / documentation, 2025–2026. LuxAlgo Price Action Concepts detects & draws SMC but does not place trades; MT5 "SMC" EAs are rules-based and single-strategy.

What the market pays today

The audience already spends here every month

LuxAlgo $35–80/mo Signal groups $70–100/mo Crypto bots $15–160/mo MT5 SMC robots $99–249 once Prop challenges $100–600 / attempt Willingness-to-pay is well established in this niche

Sources: LuxAlgo pricing; NFTEvening/Margex signal-group surveys; Koinly bot pricing; MQL5; Track360.

Sizing the opportunity (TAM → SAM → SOM)

Illustrative, bottom-up. Ranges, not promises.

TAM · $3–6B / yr retail crypto-futures traders buying tools SAM · $430–720M / yr SMC/ICT crypto traders + prop traders SOM · $7M–70M ARR 0.5–3% capture at ~$80/mo

Assumes 1–2M crypto SMC traders + slice of 2.1M funded traders; ~40% pay ~$60–120/mo. Estimates with stated assumptions.

The prop-firm channel. The funded-trader market hit ~$850M in revenue in 2026 (+45% YoY) with ~12M challenge attempts a year — and only ~7% of buyers ever get paid, usually because they break the risk rules. An automated, rule-perfect SMC engine is almost purpose-built to pass these challenges, making prop firms and their traders a natural, higher-value customer base.
Why We Win & The Business Case

The pitch, and how it makes money

The positioning writes itself: everyone else draws the setups; this one trades them. Below are the reasons it wins, and a pricing and go-to-market plan benchmarked against what this audience already pays.

01

It trades, it doesn't just draw

LuxAlgo and every SMC indicator detect and alert; they never place an order. SmartRisk closes the loop to live execution on Binance futures. That single gap is the whole pitch.

02

127 strategies vs one indicator

A library distilled from 126 SMC courses, not a single order-block drawing tool — breadth no competitor offers.

03

Confluence scoring — nobody else scores

Incumbents show you individual patterns; this ranks how many independent signals agree, so only high-conviction trades fire.

04

A physics engine no one else has

The kinematic velocity/acceleration/force timing plus a C++ execution core is an architecture no SMC indicator or MetaTrader robot can match.

05

Crypto-native, not a forex bolt-on

The rare SMC robots live on MetaTrader/forex. This is built for Binance USD-M — the largest perpetuals venue on earth.

06

It removes the #1 failure: human discretion

SMC is powerful but manual. Automation delivers the rule-consistency prop firms reward — and traders can't hold under pressure.

Pricing that mirrors the category — and reserves auto-trading for the top

Free

The hook
$0
  • Detect & draw core SMC patterns
  • Live market scanner
  • 3–5 strategies, read-only
  • Telegram alerts
  • Confluence scoring
  • Auto-execution
Most popular

Core

Serious manual traders
$59/mo
  • Everything in Free
  • Full 127-strategy library
  • Confluence scoring
  • GPU dashboard
  • Semi-auto (alert to trade)
  • Full auto-execution

Pro / Auto

Hands-free
$129/mo
  • Everything in Core
  • Full auto-execution on Binance
  • Kinematic engine + C++ core
  • Dynamic risk engine
  • Priority alerts & support

Lifetime / Prop

Own it · or white-label
$1,500+
  • Pro features, one-time
  • Prop-firm risk presets
  • Multi-account / white-label
  • API & affiliate access
  • Higher-value B2B channel

Core sits just above LuxAlgo Premium ($49.99) with far more capability; Pro is still cheaper than a single prop-challenge attempt; the lifetime and prop white-label tiers monetise the 2M+ funded-trader channel. Add annual prepay (~30% off, category norm) to fight churn.

An illustrative revenue path

Freemium funnel math — assumptions shown, not a forecast

StagePaying subscribersBlended ARPU / yrAnnual revenue
Early2,000~$900~$1.8M
Growth10,000~$960~$10M
Scale25,000~$960~$24M

Blended across $59–$129/mo tiers plus lifetime/prop. LuxAlgo's ~1.3M-follower scale shows the ceiling is far above this. Illustrative ranges, not projections.

Go-to-market, in four moves

1

Give away the best free SMC scanner

The free tier out-draws LuxAlgo's free script and pulls traders in. Screenshots of the annotated chart do the marketing.

2

Convert on confluence + the full library

Once a trader sees 127 strategies scored and ranked for them, the Core tier sells itself against a single-indicator competitor.

3

Upsell hands-free auto-trading

The Pro/Auto tier — the thing literally no competitor offers — is the premium anchor and the strongest differentiator.

4

Open the prop-firm channel

White-label to crypto prop firms with drawdown-tuned presets for concentrated, higher-value B2B contracts.

Important context. This thesis is a product and market analysis built from a direct review of the application's source code and independent research (July 2026). Market sizes, conversion rates and the revenue table are third-party estimates and illustrative assumptions, not guarantees. Parts of the system — notably the 300-feature machine-learning engine and the 127-strategy parallel hive — are documented architecture that is still largely a roadmap rather than shipping code. Trading crypto derivatives with leverage carries substantial risk of loss; the auto-trader is designed to be run in simulation before any live use. Nothing here is financial advice or a solicitation to trade.