A lightweight workflow engine for MQL5 that lets you define and run multi-step automation sequences from a simple YAML file.

---

Main Features

Define workflows in YAML

name: "my workflow"
env:
  my_var: "hello"
steps:
  - name: "RunEA"
    module: "Generic"
    with:
      fixed:
        name: "Experts\\MyEA.ex5"
        symbol: _Symbol
        timeframe: _Period
        run_flags: "DLL|AutoTrading"
        ms_sincronizacion: 500
        timeout_segundos: 300
      ea_params:
        - InpNum: ["TYPE_UINT", "${{internal.number}}"]
  - name: "PowerShellCommand"
    module: "Generic"
    with:
      run: |
        echo 'done'
      timeout_segundos: 30

Run from MQL5

#resource "my_workflow.yml" as const string yml_workflow

TSN::CWorkflow flow;
TSN::CWfCallBack* callback = new TSN::CWfCallBack();
flow.AddCallBack(callback);

// Inject runtime variables
flow.Variables().Set("internal.number", string(GetTickCount()));

if(!flow.Init(yml_workflow))
   return INIT_PARAMETERS_INCORRECT;

flow.First();

Callbacks

class CMyCallback : public TSN::IWorkflowCallback
 {
public:
  inline uint8_t FlagsCall() const override { return WORKFLOWBYLEO_CALLBACK_ALL_FLAGS; }

  bool OnWorkflowStart() override
   {
    Print("Workflow started");
    return true;
   }
  void OnWorkflowStep(int code, int step_index) override
   {
    PrintFormat("Step %d finished with code %d", step_index, code);
   }
  void OnWorkflowEnd(int code, int step_index) override
   {
    Print("Workflow ended, code: ", code);
   }
 };

Variable interpolation

Reference any variable inside the YAML using ${{key}}:

ea_params:
  - InpMagic: ["TYPE_INT",    "${{env.magic}}"]
  - InpSymbol: ["TYPE_STRING", "${{env.symbol}}"]

Built-in steps

Module	Name	Description
Generic	RunEA	Opens a chart and runs an Expert Advisor with custom parameters, waits for it to finish
Generic	PowerShellCommand	Executes a PowerShell command and waits for completion
Generic	SendMessage	Sends a message (Telegram, etc.)
Generic	File	File operations

---

Repository Structure

WorkflowsByLeo/
├── Src/
│   ├── Core/    # Core definitions: CWorkflow, CWorkflowStep, IWorkflowCallback, factory
│   ├── Steps/   # Built-in steps (RunEA, RunProcess, Messages, File)
│   └── Final/   # Orchestrator — workflow execution engine
├── Test/
│   └── First/   # Basic usage example (sync workflow: RunEA + PowerShell)
└── Images/      # Repository assets

---

License

Read Full License

By downloading or using this repository, you accept the license terms.

---

Requirements

See dependencies.json for the full list of dependencies.

Some dependencies are private or require a purchase. Check each entry's comment field for access instructions.

---

Installation

cd "C:\Users\YOUR_USER\AppData\Roaming\MetaQuotes\Terminal\YOUR_ID\MQL5\Shared Projects"
tsndep install "https://forge.mql5.io/nique_372/WorkflowsByLeo.git"

• Requires the tsndep package — available on PyPI. It automatically downloads and installs all declared dependencies.
• If any dependency is private or paid, the install will fail for that package — check dependencies.json and contact me for access.

---

Quick Start

1. Define your workflow — create a .yml file and embed it as a resource:

#resource "my_workflow.yml" as const string yml_workflow

2. Include the library:

// TSN Ecosystem file include 
#include <TSN\\WFlows\\Orquestador.mqh>
#include <TSN\\WFlows\\AllSteps.mqh>
// Or relative at your project
#include "..\\WorkflowsByLeo\\Src\\Steps\\All.mqh"

3. Instantiate, configure and run:

TSN::CWorkflow flow;
flow.Variables().Set("internal.number", string(GetTickCount()));

if(!flow.Init(yml_workflow))
   return INIT_PARAMETERS_INCORRECT;

flow.First();

4. Clean up on deinit:

void OnDeinit(const int reason)
 {
  TSN::CWorflowsFactory::Deinit(reason);
 }

See Test\First\Generic.mq5 and Test\First\test.yml for a complete working example.

---

Contact

• Platform: MQL5 Community
• Profile: https://www.mql5.com/es/users/nique_372
• Articles: https://www.mql5.com/es/users/nique_372/publications