Control AiDataTaskRunner Panel directly via MCP
Main features
Task Management
Control your backtesting queue directly from Claude:
- Add Tasks: Queue new backtests with custom symbols, dates, and timeframes
- Monitor Status: Track task progress (Processing, Queued, Pending, Ready, Failed)
- Batch Operations: Execute all pending tasks or clean completed ones
- File Management: Save/load task configurations from CSV files
Configuration Control
Manage panel settings via Claude:
- Load/Save Configurations: Save and load panel settings from text files
- Query Folders: Check data generation and task storage locations
- System Information: Access MT5 logs and system time
Natural Language Integration
Use Claude to automate your trading data pipeline:
- Query task counts and details
- Add multiple backtests in batch
- Organize and manage task configurations
- Monitor real-time execution status
Real-Time Updates
- Asynchronous operations delivered directly to chat
- Non-blocking task execution (fire-and-forget support)
- Instant status feedback from the panel
Repository Structure
AiDataTaskRunerMcp/
└── aidatataskrunner_mcp/ # Python MCP server source code
└── __init__.py
Requirements
Installation
cd "C:\Users\YOUR USER\AppData\Roaming\MetaQuotes\Terminal\YOUR ID\MQL5\Shared Projects"
tsndep install "https://forge.mql5.io/nique_372/AiDataTaskRunerMcp.git"
- For use tsndep command requerid tsndep pacakage (avaible in pypi).. This command automatically downloads all dependencies and installs all requirements from the repositories.
Quick Start (With claude ai) for final users
1. Install MCP Server
pip install aidatataskrunner-mcp
2. Create a config json
Open Common\Files
And create a file with this structure:
{
"general_config": {
"port": 9999,
"host": "localhost",
"mode": "fast_mcp"
},
"fast_mcp": {
"name": "FastMcpServer"
},
"http": {
"http_port": 8000,
"name": "HTTP Server",
"tools_namespace": "tools"
}
}
3. Configure in Claude Desktop
Add to your claude_desktop_config.json:
{
"mcpServers": {
"aidatataskrunner": {
"command": "python",
"args": ["-m",
"aidatataskrunner_mcp",
"--config",
"PATH_TO_FILE",
"--config_encodig",
"utf-8"
]
}
}
}
- PATH_TO_FILE: Path to json config file
4. Configure your MT5
Navigate to the "Tools" section >> "Options" >> "Allowed URLs for WebRequest", add a new field with the host/address you will be using (in this case "127.0.0.1"), and click Accept.
5. Connect with AiDataTaskRunner Panel
Start your AiDataTaskRunner Panel in MT5 with the AI tab enabled:
- The panel will connect to the MCP server on port 9999
- Claude will automatically detect available tools
6. Use in Claude
- Add a new EURUSD task from 2023.01.01 to 2024.01.01 on H1 timeframe, and Symbol Folder = XAUUSD, Label id = 0
- Give me the total number of tasks I have
- Clean all task
- Run all task
Claude will automatically translate this to the appropriate MCP call.
Available Tools
Task Management
aidatataskrunner_add_task — Add a new backtest task
aidatataskrunner_get_task_total — Get total task count
aidatataskrunner_get_task_by_index — Get task details by index
aidatataskrunner_get_task_status — Get task execution status
aidatataskrunner_execute_all_tasks — Execute all pending tasks
aidatataskrunner_clean_all_tasks — Clean completed tasks
File Operations
aidatataskrunner_save_tasks_to_file — Export tasks to CSV
aidatataskrunner_load_tasks_from_file — Import tasks from CSV
aidatataskrunner_save_config — Save panel configuration
aidatataskrunner_load_config — Load panel configuration
Information Queries
aidatataskrunner_get_main_folder — Get main data folder path
aidatataskrunner_get_task_folder — Get tasks folder path
aidatataskrunner_is_in_commonfolder — Check if using common MT5 folder
License
Read Full License
By downloading or using this repository, you accept the license terms.
Documentation
Full documentation is available in the main repository:
Contact
Copyright © 2026 Nique-Leo.