Техническое задание
This is the Requirements Specification for your project, formatted to be clear, concise, and structured. This document will allow a professional MQL5/Python developer to accurately estimate the complexity, time, and cost of building your IATS.
IATS Project: Expert Advisor (MQL5) & Python Hub Specification
Job Title: AI-Validated Algorithmic Execution System Developer (MQL5/Python)
Project Goal: Build the high-speed execution engine and orchestration layer for a risk-first automated trading platform.
I. System Architecture & Environment
| Component | Responsibility | Technology | Requirements |
|---|---|---|---|
| Execution Engine | Trade Management, Data Feed, Auditing. | MQL5 (MT5) | Must be compiled for MetaTrader 5 (64-bit). |
| Orchestration Hub | AI Calls, Risk Calculation, Dashboard Server. | Python 3.10+ (with requests, Streamlit, SQLite) | Must run reliably 24/7 on a Windows VPS. |
| Communication | Bidirectional link between Python and MQL5. | Local HTTP/WebSockets (or file-based bridge) | Must ensure near-instantaneous (low-latency) command execution. |
| Risk Boundary | Max Risk Per Trade. | 1.0% of current account equity. | Non-negotiable hard limit. |
II. Module 1: MQL5 Expert Advisor (The Engine)
The MQL5 EA must perform three core functions: Execution, Management, and Data Fetching.
A. Execution and Lot Sizing
* Instruction Input: The EA must constantly listen for and accept a single JSON command payload from the Python Hub, containing: Symbol, Type, Entry, SL, TP1, TP2, TP3, and the three pre-calculated Lot Sizes (\mathbf{40\%/30\%/30\%}).
* Order Placement: Upon receiving a command, the EA must immediately open three separate market orders (Orders A, B, C) simultaneously for the target symbol, ensuring all orders share the same Initial Stop Loss (SL) but have unique Take Profit (TP) levels (TP1, TP2, TP3, respectively).
* Unique Tracking: All three orders must be placed with a shared, custom Magic Number for unit tracking.
B. Mandatory Trade Manager (40%/30%/30% Protocol)
This logic must be implemented inside the OnTick() loop and use status flags to ensure rules are only executed once per trigger.
| Rule | Trigger Condition (Example: BUY Trade) | Required Action (Using PositionClosePartial and PositionModify) |
|---|---|---|
| Rule 1: Risk Elimination | Current Price (Bid) >= TP1 | 1. Close Order A (40% volume). 2. IMMEDIATELY Modify SL for the remaining \mathbf{60\%} (Orders B & C) to the Initial Entry Price (Breakeven). |
| Rule 2: Profit Securing | Current Price (Bid) >= TP2 and Rule 1 is complete. | Close Order B (30% volume). |
| Rule 3: Final Close | Current Price (Bid) >= TP3 and Rule 2 is complete. | Close Order C (30% volume), concluding the trade cycle. |
C. Data Feed and Auditing
* Market Context Feed: The EA must calculate and package two pieces of data to send to the Python Hub upon request:
* Trend Bias: Current price position relative to the 200-period EMA on the H4 timeframe (ABOVE or BELOW).
* Volatility Check: Current Bid/Ask Spread in pips.
* Audit Data Capture: Upon complete closure of the trade unit (Orders A, B, C are closed), the EA must calculate and log:
* Final P/L (Profit/Loss).
* Trade Duration.
* Maximum Intrade Drawdown (MID): The largest unfavorable movement suffered by the position from the entry price before final closure.
III. Module 3: Python Orchestration Hub
The Python Hub manages external communication, the AI pipeline, and the final risk gate.
A. Core Risk and Lot Sizing
* Maximum Risk Enforcement: The Hub must calculate the Total Lot Size based strictly on the current account balance and the \mathbf{1.0\%} risk limit, using the \text{SL} distance extracted by Agent 1.
* Volume Allocation: The calculated Total Lot Size must be split into 40\%, 30\%, and 30\% components before being sent to the MQL5 EA.
B. Three-Stage Gemini AI Pipeline
The Hub must execute three sequential, independent calls to the Gemini 2.5 Flash API (TEXT_API_URL) for every incoming signal. The pipeline halts on any Veto signal.
| Agent | Input Data | Output Veto Rule |
|---|---|---|
| Agent 1: Translator | Raw Telegram Text | VETO: If JSON output is malformed or critical data (SL, TP) is missing. |
| Agent 2: Auditor | Raw Telegram Text | VETO: If Confidence Score is below 80 (threshold must be dynamically adjustable via the Streamlit Dashboard). |
| Agent 3: Strategist | Clean JSON from Agent 1 + Market Context Data from MQL5 | VETO: If the AI determines the trade violates a strategic rule (e.g., trading against the H4 200 EMA, or Spread > 2.0 pips). |
C. Dashboard and State Management
* Streamlit Integration: Host the Streamlit web application on the VPS, dynamically reading and writing configuration variables (the three AI Prompts and Veto thresholds) to the SQLite database.
* Live Data Integration: Pull and display current equity, P/L, and the logged MID/Audit data from the SQLite database onto the dashboard.
* Gemini Dashboard Features: Implement the Auto-Refine Prompt (Text Generation API) and the TTS Weekly Recap (TTS API).
IV. Required Expertise & Timeframe
| Requirement | Level | Note |
|---|---|---|
| MQL5 Development | Expert | Deep experience required with complex order modification (PositionModify) and partial closures (PositionClosePartial). |
| Python Development | Expert | Required for multi-API orchestration (Gemini, Telegram), database management, and web framework (Streamlit) deployment. |
| Development Time | 8 Weeks | This is the minimum required timeline for clean coding and rigorous cross-platform testing of the complex bi-directional communication layer and trade management logic. |
Откликнулись
1
Оценка
Проекты
30
20%
Арбитраж
8
63%
/
13%
Просрочено
9
30%
Свободен
2
Оценка
Проекты
47
85%
Арбитраж
1
0%
/
100%
Просрочено
9
19%
Свободен
3
Оценка
Проекты
497
19%
Арбитраж
33
42%
/
30%
Просрочено
32
6%
Занят
4
Оценка
Проекты
196
48%
Арбитраж
3
33%
/
33%
Просрочено
1
1%
Работает
5
Оценка
Проекты
1
0%
Арбитраж
0
Просрочено
0
Работает
6
Оценка
Проекты
12
33%
Арбитраж
1
0%
/
100%
Просрочено
1
8%
Свободен
7
Оценка
Проекты
1
0%
Арбитраж
0
Просрочено
0
Свободен
Опубликовал: 2 статьи
8
Оценка
Проекты
83
64%
Арбитраж
5
100%
/
0%
Просрочено
9
11%
Свободен
9
Оценка
Проекты
34
24%
Арбитраж
4
0%
/
50%
Просрочено
2
6%
Работает
10
Оценка
Проекты
469
39%
Арбитраж
100
41%
/
23%
Просрочено
77
16%
Загружен
Опубликовал: 2 примера
11
Оценка
Проекты
20
0%
Арбитраж
10
0%
/
80%
Просрочено
6
30%
Свободен
12
Оценка
Проекты
7
0%
Арбитраж
3
33%
/
33%
Просрочено
3
43%
Занят
13
Оценка
Проекты
1
0%
Арбитраж
1
100%
/
0%
Просрочено
0
Свободен
Похожие заказы
Ninjatrader indicator
30+ USD
I have an issue with my ninja script and i would like you to help me straighten things I wanted to create an indicator and i have the source code already but i am getting compiling errors on my NinjaTrader And i tried fixing the error it still same I sent 3 images here for you to understand the errors and i would like to ask if you can help me fix it so i can go ahead and compile my source code. Thanks
Nijatrader indicator
30+ USD
Good day, I would like to build an automated trading system for Ninjatrader using 2 MACD, a Supertrend, and a moving average indicator. I want the option to adjust the indicator settings, the ability to trade at three different times, and the option to receive alerts. I want to get an idea of what that will cost me. It will enter trades on all blue take one contract out at a fixed point, move the stop to break even
"Hello! I am an experienced programmer specializing in automated trading software for MetaTrader 4 (MQL4) and MetaTrader 5 (MQL5). My goal is to help traders turn their manual strategies into fully automated robots (Expert Advisors) and custom indicators. My services include: Developing Expert Advisors (EA) from scratch based on your strategy. Creating Custom Indicators and Scripts. Modifying existing EAs (adding
I have a simple strategy that need coding on tradingview Strategy using high low at seleted time and when breakout to entry buy sell. Everything will be explained on private
Enhance current EA
200+ USD
I am looking for an experienced MQL5 developer to modify and enhance my existing Expert Advisor, "Gold Levels Trader". The current version has a low win rate (~30%) and issues with ATR-based Stop Loss execution. I want to replace the current "pips drop/rise" logic with Fibonacci Retracement levels for entry signals, implement a Daily Drawdown Limit , and add Pending Orders functionality
I want developer who know how to create bot which immediately transfer specific crypto coin deposit to one crypto address to another specific address in just a second,, if you know about this then only comment on this post
So the things we need in algorithm of mql5 language EA in mt5 1. Depending on timeframe it can recognise the previous swing high and low 2. Timeframe is 5m,15m,1h,4h 3. It can recognise the basic Market bias that is market is bullish or bearish we can identify using (ema,rsi,basic smc bias,ict bias structure mapping) or use anything to find bias structure 4. EA should have option to change timeframe and change risk
require the development of a high-speed, fully automated trading Expert Advisor (EA) for MetaTrader 5 , optimized for live trading on both Deriv and Exness . The EA must be designed for fast execution, low latency, and reliability on real-money accounts , with full compatibility across broker-specific contract specifications, tick sizes, tick values, pricing formats, and volume rules. It should automatically detect
# Copyright 2025, MetaQuotes Ltd. # https://www.mql5.com import importlib . util mt5_spec = importlib . util . find_spec ( "MetaTrader5" ) if mt5_spec is None : mt5 = None else : import MetaTrader5 as mt5 # pyright: ignore[reportMissingImports] def main (): if mt5 is None : print ( "MetaTrader5 module not available" ) return print ( 'Python executable:' , __import__ ( 'sys' ).executable) ok = mt5
Expert Advisor – Automated Forex Trading System (MT5)
500 - 3000 USD
Scope of Work Develop 15 individual Expert Advisors (MT4 and/or MT5) Each EA will have: Unique trade logic and execution rules Configurable inputs (risk %, lot size, TP/SL, filters, sessions, etc.) Clean, modular, well-commented code Error-free compilation Backtest-ready functionality Core Features (Across EAs) Market & pending order execution Risk-based position sizing Time/session filters Trade limits (per day /
Информация о проекте
Бюджет
100+ USD
Сроки выполнения
от 1 до 20 дн.