🧮 Multi-timeframe matrix · 15m → 1D alignment — every frame votes trend + RSI + MACD + EMA posture · STACKED = all frames agree · HTF-LEAD + LTF-pullback = the elite dip/rip context
reading frames…
📣 Social radar · crowd attention + sentiment — LunarCrush galaxy/altrank when keyed · CoinGecko trending + Fear & Greed + news-wire tone, keyless · heat = attention, dir = directional read — a coin can trend on FUD
reading the crowd…
🔥 Signal pressure · how close the engine is to firing
near-miss candidates, next 10 below the cutoff—
🧪 Evidence / validation · prospective results only — emitted signals are measured forward, traded or not · the number that matters is whether predictions precede outcomes
💡 Hypothesis engine · what the evidence currently supports — claims escalate with sample: UNTESTED <10 · EARLY <30 · SUGGESTIVE <100 · then SUPPORTED/REFUTED
🧬 Einstein lab · Head of R&D — IT/DevOps — autonomous research: per-factor alpha lift, pipeline diagnostics, structural clustering, improvement proposals · intelligence-surface authority, execution stays GOD-gated
lab initialising…
📰 Intelligence wire · live headlines — public RSS, asset-tagged · context only, never a score input
🏆 Ranked signals · momentum confluence — click a row for detail · positions open only at score ≥70 (≥75 volatile) + strategy must fit the market type — wrong-family setups stand down
Asset
Side
Grade
Score
24h
RSI 1h
Price
Target
48h
📈 Top movers / laggards · 24h
Movers
Laggards
💰 Funding extremes · 8h rate
Longs pay
Shorts pay
🔊 Volume leaders · 24h turnover
📜 Trade journal — Taken: real exchange round-trips with fees · Emitted: every signal graded against the tape it predicted (click a row for before/after chartwork)TakenEmitted
🚀 Live execution · real Bitget account — exchange state is ground truth
🛠 Trade setups · operator command — market / bounce / break-&-retest entries · risk-% sizing · multi-tier take-profits · stop-market SL · break-even after TP1 — every fill rides the same exchange-side rails as the engine
no armed setups — build one below, send /setup in Telegram, or fire a TradingView alert
⚙️ BUILD A SETUP — the risk card computes live as you type · arming writes straight into the executor queue (webhook) or copies a /setup command for Telegram
TAKE-PROFIT LADDER — pct of position; unallocated remainder rides the stop as the moon bag
webhook intake: POST {this host}/tv-hook — TradingView alert body template lives in the box under scripts/tv-hook.py. Paper-safe: fills land in demo mode under the same rails as every entry.
🧮 POSITION SIZE CALCULATOR — identical math to the executor's risk-% path: what you see is what it sizes
TP TARGETS
🏦 Wealth vault · profits locked forever — 50% of every winning close sweeps out of the tradable book — the engine cannot touch it, not ever
⚖️ Payoff engine · empirical edge × Monte Carlo — fitted entry edge gates and scales every entry · ladder profile re-picked by billion-roll evidence each 30min
✅ Integrity · CRC32 tamper-evident record — every artifact checksums into the manifest on write, verifies on read, audits on a 15min loop
👁️ GOD · the overseer — audits every system invariant each cycle · pure read, never mutates, cannot hide
💸 Funding radar · Bitget perp funding arb scan — delta-neutral harvest candidates + carry exposure of open trades
Coin
Funding 8h
Annualized
Basis
Arb play
Breakeven
📊 Benchmark · this engine vs the paid-signal field — real Bitget account equity vs passive benchmarks, framed by Sloggett's published framework
vs boring benchmarks — does the intelligence add value beyond doing something trivial?
— sentinel vs - - BTC hold · % since clock start
Paid signal services (Alpha Signals included) publish entries and marketing — not an auditable record, R-multiples, or SQN. This engine's entire record is public: every signal is timestamped and archived, and every fill comes straight from the Bitget exchange — committed to git and independently verifiable against Bitget candles. The performance benchmark gets earned by the real account.