Skip to main content

Weather Connector Example

This example demonstrates a complete weather connector that fetches current conditions and forecasts using the Open-Meteo API.

Structure​

  • connector.yaml - Connector definition with two actions
  • workflows.yaml - Multiple workflows example (daily report, alerts, health check)
  • service.py - HTTP service implementation (Python/FastAPI)

Actions​

get_current_weather​

Fetches current weather conditions for a location.

Input:

{
  "location": "San Francisco"
}

Output:

{
  "location": "San Francisco",
  "latitude": 37.7749,
  "longitude": -122.4194,
  "temperature": 18.5,
  "windspeed": 12.3,
  "wind_direction": 280,
  "weather_code": 2,
  "time": "2025-10-07T12:00:00Z",
  "status": "success"
}

get_forecast​

Gets 7-day weather forecast.

Input:

{
  "location": "New York",
  "days": 7
}

Running Locally​

Install Connector​

# Load connector into SDK
smartify add connector:community/weather

# Or install from local file
smartify add ./connector.yaml

Execute Workflows​

Python SDK:

from openworkflow import OpenWorkflow, Workflows

openworkflow = OpenWorkflow(execution_mode="local")
openworkflow.register_connector("connector.yaml")

# Load all workflows
workflows = Workflows.from_file("workflows.yaml")

# Execute specific workflow
report_result = workflows.execute("Daily Weather Report", inputs={
    "locations": ["San Francisco", "Seattle"]
})

# Execute health check
health_result = workflows.execute("Service Health Check")

# Or execute all workflows
results = workflows.execute_all()

CLI:

# Execute specific workflow
smartify workflow run workflows.yaml --workflow "Daily Weather Report" --input locations='["Boston","NYC"]'

# List available workflows
smartify workflow list workflows.yaml

# Execute all workflows
smartify workflow run workflows.yaml --all

Run as HTTP Service​

# Install dependencies
pip install fastapi uvicorn requests

# Start service
uvicorn service:app --host 0.0.0.0 --port 8000

# Test endpoint
curl -X POST http://localhost:8000/execute \
  -H "Content-Type: application/json" \
  -d '{"location": "San Francisco"}'

Deploying​

Docker​

# Build image
docker build -t weather-plugin .

# Run container
docker run -d -p 8000:8000 \
  -e MCP_ROUTER_URL=http://mcp-router:8000 \
  weather-plugin

Kubernetes​

kubectl apply -f k8s/deployment.yaml

OpenWorkflow Cloud​

# Publish to registry
smartify publish connector.yaml

# Deploy workflow
smartify workflow deploy workflows.yaml

Testing​

# Validate connector schema
smartify validate connector.yaml

# Test workflow
smartify workflow test workflow.yaml --input locations='["Boston","Seattle"]'

Implementation Notes​

Error Handling​

The connector implements retry logic with exponential backoff for external API calls:

  • Geocoding API (Nominatim)
  • Weather API (Open-Meteo)

Rate Limiting​

Open-Meteo API limits:

  • 10,000 requests/day for free tier
  • Consider caching results for same location within time window

Weather Codes​

WMO weather codes:

  • 0: Clear sky
  • 1-3: Partly cloudy
  • 45, 48: Fog
  • 51-67: Rain
  • 71-77: Snow
  • 80-99: Thunderstorm

See: https://open-meteo.com/en/docs

Extending​

Add New Actions​

Add to connector.yaml:

actions:
  - name: get_air_quality
    description: Get air quality index for a location
    # ... action definition

Custom Handler​

Use SDK function instead of HTTP:

from openworkflow import action

@action("weather.get_current_weather")
def get_weather(location: str) -> dict:
    # Custom implementation
    return {"temperature": 20, "location": location}

References​