Deterministic intelligence
Observes and decides
Multi-timeframe structure, trend, momentum, volatility, regime, Strategy Brain, and BUY / SELL / WAIT run without a generative model.
AI Intelligence
Intelligence you can inspect. Authority that stays deterministic.
GoldMind separates three jobs clearly: market and strategy logic form the decision, optional AI explains the evidence, and deterministic safety controls whether execution is allowed.
Visual Coach + AI Tutor
Ask why a setup qualifies—or why GoldMind stays out.
The Visual Coach places chart evidence beside structured reasoning. The optional AI Tutor translates bounded context into customer language while the original decision, risk state, and execution permission remain unchanged.
- Chart marks show proposed or executed ROBOT levels
- WAIT and blocked states receive explanations too
- Provider credentials remain outside the EA/cBot
- Deterministic engines continue when AI is unavailable
Visual Coach · Chart evidence
Explanation onlySTRUCTURE · BULLISHTREND · ALIGNEDMOMENTUM · BUILDINGVOLATILITY · NORMAL
ENTRY
Proposed
SL
Protected
TP1 / TP2
Planned
Why this setup?
Context is constructive. Entry is not confirmed.
◆
H4 / H1
Directional structure agrees
◆
M15
Momentum is improving
◇
M5
Trigger candle is incomplete
◇
Safety
No execution path while WAIT
Optional AI Tutor
Explains the bounded evidence in customer language. It cannot change the decision, risk limits, or execution permission.
The authority map
Explanation sits beside the pipeline—not above it.
Optional AI can summarize evidence at several points. It cannot inject a new decision, bypass WAIT, alter sizing, arm Auto, or approve broker I/O.
Strategy system
Evidence becomes a proposal. Safety decides what may continue.
01◫
Market Data
H4 · H1 · M15 · M5
02⌁
Evidence
Structure · trend · volatility
03◆
Strategy Brain
Selects a strategy family
04◇
BUY / SELL / WAIT
Proposes a decision
BUYSELLWAIT
05◈
Risk Engine
Authorizes or blocks
06↗
Execution
Mode + final recheck
07◎
Position Monitor
ROBOT lifecycle only
AI / Strategy: proposes and explains.
Risk / Safety: permits or blocks.
WAIT: remains a complete, valid result.
Optional AI explanation
Explains and teaches
WHY_SIGNAL, trade-analysis, tutor, news-analysis, and review tasks receive bounded context. Their output cannot place or authorize orders.
Offline research
Tests and reviews
Journal evidence, OOS, walk-forward, stress, and Monte Carlo support research. They do not autonomously mutate live trading logic.
AI / Strategy proposal · non-authoritative
XAUUSD
Normal · BUY candidate
ENTRY
2345.20
SL
2338.40
TP
2360.10
Deterministic safety authority
◈Data qualityFRESH
News windowCLEAR
Spread / marginPASS
Daily loss controlCLEAR
Weekly loss controlCLEAR
Final recheckAPPROVE
ORDER PATHAPPROVED
Hard safety boundary
Safety is code, not persuasion.
Data freshness, news protection, spread, margin, per-trade risk, daily and weekly controls, portfolio exposure, terminal permissions, kill switch, ownership, and the final recheck remain authoritative.
The AI layer cannot change these checks, their thresholds, or their outcome.
Read safety documentation