Quant AI Agents
- Utilità
- Ho Tuan Thang
- Versione: 1.0
- Attivazioni: 10
WANT THE SAME RESULTS AS MY LIVE SIGNAL?
Use the exact same brokers I do: IC MARKETS, IC TRADING . Unlike the centralized stock market, Forex has no single, unified price feed. Every broker sources liquidity from different providers, creating unique data streams. Other brokers can only achieve trading performance equivalent to 60-80%.
Forex EA Trading Channel on MQL5: Join my MQL5 channel to update the latest news from me. My community of over 15,000 members on MQL5.
ONLY 3 COPIES OUT OF 10 LEFT AT $199!
After that, the price will be raised to $299.
EA will be sold in limited quantities to ensure the rights of all customers who have purchased.
USER GUIDE & SETUP FOR QUANT AI AGENTS
Quant AI Agents is not just another Expert Advisor — it is a multi-agent AI trading framework that turns natural-language strategy prompts into live, fully-managed orders. Where every other EA on this market sells you ONE hard-coded strategy, Quant AI Agents ships as a strategy-engine: a verified built-in preset for XAU/USD plus a public Prompt Library of professionally written quant prompts (Donchian Turtle, Connors RSI2, IBS Equity, Turtle Soup, NR7, ACD Gold NY and more), and the ability to load your own unlimited custom strategies from a single text file. One EA — hundreds of strategies — one decision pipeline.
Under the hood, every trading decision is debated by five specialised AI agents — Technical Analyst, Bull Analyst, Bear Analyst, Research Manager and Portfolio Manager — each with its own role, prompt, and veto power. The Bull and Bear actively argue against each other across multiple debate rounds; the Research Manager weighs the evidence; the Portfolio Manager has the final word and translates conviction into position size, stop and exit plan. This adversarial design eliminates the single-model blind spot that plagues every "ChatGPT EA" you have seen so far.
Crucially, the system remembers and learns. Every closed trade — the prompt that produced it, the agents' reasoning, the entry context and the realised P&L — is written to a per-instance memory file. Before each new analysis, agents re-read recent decisions, per-symbol reflections and global lessons learned. Losing prompts get downgraded automatically; recurring failure patterns surface as warnings to the Bear analyst. The bot literally gets sharper the longer it runs on your account.
THREE CORE STRENGTHS
1. A Strategy Engine, Not a Frozen EA. Most EAs lock you into a single thesis — if the market regime changes, the EA dies with it. Quant AI Agents lets you switch strategies the way a professional trader rotates playbooks: load a mean-reversion prompt this week, swap to a breakout prompt next week, or run several presets across different charts simultaneously. The verified built-in preset is ready out-of-the-box; the public Prompt Library covers eight battle-tested quant archetypes; and the open prompt-file format means you can deploy any strategy you (or your community) can describe in plain English.
2. Five-Agent Debate Pipeline. A single AI call is a guess. Five agents arguing is a process.
- Technical Analyst — reads the chart, indicators and price action.
- Bull Analyst — constructs the strongest possible case to enter.
- Bear Analyst — constructs the strongest possible case to stay out or invalidate.
- Research Manager — arbitrates the debate, scores conviction.
- Portfolio Manager — sizes the position, sets stops, applies risk overlays, places the order.
Configurable debate depth lets you tune speed-vs-thoroughness; in deep mode the Bull and Bear iterate multiple rounds until the Research Manager is satisfied that the case has been stress-tested.
3. Persistent Memory & Self-Reflection. After every closed trade, the system runs a reflection pass: What did the agents see? What was the outcome? Which assumption broke? Reflections are stored per-symbol and aggregated globally. On the next decision cycle, the agents read this history before opening their analysis — so a strategy that just bled on three NFP releases will see Bear weight that pattern more aggressively next time. This closes the loop between live performance and prompt selection without any human intervention.
BUILT-IN INSTITUTIONAL-GRADE RISK MANAGEMENT
The decision pipeline is wrapped in a multi-layer risk governor that runs independently of the AI — even if the agents' conviction is high, hard limits cannot be overridden:
- Daily / weekly / monthly drawdown caps with auto-lockout on breach
- Equity-floor protection vs initial deposit
- Loss-streak damper that auto-shrinks risk after consecutive losses
- Daily profit-lock to protect green days
- Margin-level guard, aggregate-risk cap, post-loss cooldown
- High-impact news blackout (NFP, FOMC, CPI …) with configurable currencies and importance threshold
- Friday cut-off and Sunday open blackout to skip illiquid windows
- Volatility-spike auto-close, time-based break-even, trailing stop, partial-close on R-multiple targets
- Optional position averaging with hard aggregate-volume cap and group-kill drawdown
NO MARTINGALE. NO GRID. NO RECOVERY ZONES. Every position is sized from a stop distance and a risk percentage; losing trades are accepted, not doubled.
SUPPORTED AI MODELS
Plug in any of the major API providers via a single API-key field — the same key powers all five agents:
- OpenAI GPT family (GPT-4o, GPT-4o-mini and newer)
- DeepSeek (cost-efficient, all five agents on one key)
- Anthropic Claude family
- Google Gemini
- Custom OpenAI-compatible endpoints (self-hosted, Azure, OpenRouter …)
You choose the model, you control the cost, you keep your key private — nothing routes through my server.
ADVANCED FEATURES
- Multi-agent debate pipeline with configurable depth (fast / standard / full)
- Public Prompt Library + unlimited custom prompts via plain-text file
- Per-instance trade memory with automatic reflection after every close
- Multi-layer hard risk governance, independent of AI conviction
- Volatility-adaptive exit engine (regime-aware trailing, Chandelier, MFE time-stop, adaptive vol-spike)
- High-impact news blackout with per-currency control
- Mass-concurrency safe — run multiple instances on the same terminal across symbols / timeframes / strategies
- Built-in trading panel with one-click manual override
…
| The verified built-in preset has been rigorously backtested against multi-year tick history with realistic spread, slippage and commission. Live evaluation across multiple brokers and account sizes is ongoing — results will be published on the official MQL5 signal page once the supervised period is complete. |
- EA CONFIGURATION:
| Symbol | XAU/USD (built-in preset). Other symbols supported via custom prompts. |
| Timeframe | H4 (recommended for built-in preset). Other timeframes supported via prompts. |
| Test From | Multi-year tick history |
| Settings | Default optimised parameters — load and run. |
| Brokers | True ECN Brokers: IC Markets, IC Trading Unlike the centralized stock market, Forex has no single, unified price feed. Every broker sources liquidity from different providers, creating unique data streams. Other brokers can only achieve trading performance equivalent to 60-80%. |
| AI Model | OpenAI / DeepSeek / Claude / Gemini / Custom endpoint — one API key powers all five agents. |
| Minimum Deposit | $500 / 0.01 lot |
| Recommend Deposit | $1000 / 0.01 lot (target ≤10% drawdown) |
| Advanced Features | - Five-agent debate pipeline (Technical / Bull / Bear / Research / Portfolio) - Persistent per-instance trade memory with automatic post-trade reflection - Multi-layer hard risk governance independent of AI conviction - Regime-adaptive exit engine, news blackout, session & day-of-week filters - Mass-concurrency safe across multiple charts / symbols / strategies |
Security Notice:
|
