AI-powered trading strategy generation using n8n workflows and Python ML engine.

• Python 3.8+
• n8n instance (self-hosted or n8n.cloud)
• n8n API key

1. Clone the repository

   git clone https://github.com/yourusername/n8n_github.git
   cd n8n_github

2. Create Python virtual environment

   python3 -m venv venv
   source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install dependencies

   pip install -r scripts/requirements.txt

4. Configure n8n API credentials

   cp .env.example .env
   # Edit .env and add your n8n credentials

1. Log in to your n8n instance

   • If using n8n.cloud: https://app.n8n.cloud
   • If self-hosted: http://localhost:5678 (default)

2. Generate API Key

   • Click your avatar (top-right)
   • Select "Account Settings"
   • Navigate to "API Tokens"
   • Click "Create New Token"
   • Copy the token

3. Update .env file

   N8N_BASE_URL=https://yourname.n8n.cloud
   N8N_API_KEY=your_api_key_here
   N8N_API_VERSION=v1
   N8N_TIMEOUT=30
   N8N_VERIFY_SSL=true

Test your n8n API configuration:

python scripts/test_n8n_connection.py

Expected output:

1️⃣  Testing .env file... ✅
2️⃣  Testing configuration... ✅
3️⃣  Testing API client creation... ✅
4️⃣  Testing API connection... ✅
5️⃣  Testing workflow listing... ✅

✅ All tests passed! n8n API is ready to use.

from scripts.n8n_integration import N8nIntegration
from dotenv import load_dotenv

load_dotenv()

integration = N8nIntegration()
if integration.connect():
    integration.print_workflows_table()

from scripts.n8n_integration import N8nIntegration
from dotenv import load_dotenv

load_dotenv()

integration = N8nIntegration()
integration.connect()

# Execute workflow and wait for completion
result = integration.execute(
    workflow_id='your_workflow_id',
    input_data={'symbol': 'BTC', 'timeframe': '1h'},
    wait=True,
    max_wait=300
)

print(f"Status: {result.get('status')}")
print(f"Data: {result.get('data')}")

from scripts.execute_trading_workflow import TradingWorkflowExecutor
from dotenv import load_dotenv

load_dotenv()

executor = TradingWorkflowExecutor()
executor.connect()

# Find input CSV
input_csv = executor.find_input_csv('BTC_USDT')

# Execute workflow
execution = executor.execute_workflow(
    'DailyFeatureSelection',
    input_csv=input_csv,
    wait=True
)

# Save results
executor.save_execution_result(execution, 'trading_result.json')

The repository includes a small agent that periodically triggers a configured n8n workflow (e.g. a strategy generator), waits for it to complete, and saves results.

Configuration (add to .env):

• WORKFLOW_NAME or WORKFLOW_ID — workflow to trigger (default: StrategyGenerator)
• AGENT_INTERVAL — seconds between runs (default: 3600)
• DRY_RUN — set to true to run the agent locally without contacting n8n (useful for testing)
• UPLOAD_TO_GCS, GCS_BUCKET — optional: upload saved results to GCS
• BQ_TABLE — optional BigQuery table project.dataset.table to load tabular results

Run the agent (foreground):

source .venv/bin/activate
python scripts/strategy_agent.py

Run in dry mode (no external credentials required):

export DRY_RUN=true
python scripts/strategy_agent.py

Where results are saved:

• Saved result files go to the reports/ directory by default.
• A sample result is included at reports/sample_strategy.json.

If you set DRY_RUN=true the agent will generate a fake execution result and store it locally so you can test the full pipeline without network or credential configuration.

n8n_github/
├── scripts/
│   ├── n8n_api_client.py          # Low-level API client
│   ├── n8n_integration.py         # High-level integration
│   ├── execute_trading_workflow.py # Trading workflow executor
│   ├── example_n8n_usage.py       # Full usage examples
│   ├── test_n8n_connection.py     # Connection tests
│   └── requirements.txt           # Python dependencies
├── data/                          # Input CSV data
├── reports/                       # Output results & analysis
├── .env                           # Configuration (ignored by git)
├── .env.example                   # Configuration template
└── README.md                      # This file

• test_connection() - Verify API connectivity
• list_workflows() - Get all workflows
• get_workflow(workflow_id) - Get workflow details
• execute_workflow(workflow_id, data) - Execute workflow
• get_execution(execution_id) - Get execution status
• list_executions(workflow_id) - List executions
• wait_for_execution(execution_id) - Wait for completion

• connect() - Test connection
• get_workflows() - Get workflows with caching
• find_workflow(name_or_id) - Search workflows
• execute() - Execute with defaults
• get_execution_status() - Get status
• list_recent_executions() - List recent runs

• connect() - Connect to n8n
• list_available_workflows() - Display workflows
• find_input_csv() - Locate CSV files
• read_csv_data() - Read CSV data
• execute_workflow() - Execute with CSV
• save_execution_result() - Save results
• process_workflow_results() - Format results

1. Check N8N_BASE_URL is correct

   curl https://yourname.n8n.cloud/api/v1/me -H "X-N8N-API-KEY: your_key"

2. Verify API key is valid

   • Regenerate in n8n → Account Settings → API Tokens

3. Check firewall/network access

   • Ensure n8n instance is accessible

1. List available workflows

   python scripts/n8n_integration.py

2. Use exact workflow ID from list

   • Format: "123456" or "abcdef123456"

Increase timeout in .env:

N8N_TIMEOUT=60
N8N_MAX_WAIT_SECONDS=600

• n8n Documentation
• n8n API Reference
• Trading Strategy Guide
• Quick Start (60 min)

MIT

1. Create a feature branch
2. Make changes
3. Test with python scripts/test_n8n_connection.py
4. Submit a pull request