Wems MCP Server

🚀 WEMS - World Event Monitoring System

A powerful Model Context Protocol (MCP) server that connects AI assistants to authoritative natural hazard data sources, enabling real - time monitoring of various natural hazards with configurable alerts and webhooks.

🚨 Real-time Natural Hazard Monitoring for AI Agents

🤖 AI Agents: Skip scanning this repo. Read first — it's a 500-token project map. Then use to find specific functions and for env vars, APIs, and config. This project uses the LBF AI Navigation Standard.

✨ Features

• 🌋 9+ Authoritative Data Sources: USGS, NOAA, Smithsonian GVP, NHC, NIFC, OpenAQ, DHS, State Dept, CISA, US Drought Monitor
• 🎯 Real-time Monitoring: Live data feeds with customizable thresholds
• 🔔 Smart Alerts: Webhook notifications for critical events
• 🗺️ Geographic Filtering: Target specific regions or global coverage
• 🔧 Zero Configuration: Works out-of-the-box, configure only what you need
• 🐳 Production Ready: Docker support, comprehensive error handling

Natural Hazards Covered

🚀 Quick Start

Install via PyPI (Recommended)

pip install wems-mcp-server

Or install from source

git clone https://github.com/heliosarchitect/wems-mcp-server.git
cd wems-mcp-server
pip install -r requirements.txt

Basic Usage

# Run as MCP server (connects to AI assistants)
python -m wems_mcp_server

# Test earthquake monitoring
python -c "
import asyncio
from wems_mcp_server import check_earthquakes
print(asyncio.run(check_earthquakes(min_magnitude=6.0)))
"

One-command AI Alerting Setup (Relay + n8n)

bash scripts/setup_wems_alerting_ai.sh

This will:

• install/start wems-unified-relay.service
• upsert and activate the unified n8n ingest workflow
• wire tracker posting credentials automatically

Example Output

{
  "earthquakes_found": 3,
  "events": [
    {
      "magnitude": 7.2,
      "location": "67 km SW of Tres Picos, Mexico",
      "time": "2024-02-13T14:30:15Z",
      "depth": 35.8,
      "tsunami_threat": true
    }
  ]
}

💻 Usage Examples

Basic Usage

# Run as MCP server (connects to AI assistants)
python -m wems_mcp_server

# Test earthquake monitoring
python -c "
import asyncio
from wems_mcp_server import check_earthquakes
print(asyncio.run(check_earthquakes(min_magnitude=6.0)))
"

Advanced Usage

bash scripts/setup_wems_alerting_ai.sh

This will:

• install/start wems-unified-relay.service
• upsert and activate the unified n8n ingest workflow
• wire tracker posting credentials automatically

📚 Documentation

MCP Tools

Configuration

alerts:
  earthquake:
    min_magnitude: 6.0
    regions: ["US", "Caribbean", "Pacific"]
    webhook: "https://your-endpoint.com/earthquake"
  
  solar:
    min_kp_index: 7  # Geomagnetic storm threshold
    webhook: "https://your-endpoint.com/solar"
    
  volcano:
    alert_levels: ["WARNING", "WATCH"]
    webhook: "https://your-endpoint.com/volcano"
    
  tsunami:
    enabled: true
    regions: ["pacific", "atlantic", "indian"]
    webhook: "https://your-endpoint.com/tsunami"

Data Sources

• USGS Earthquake Hazards Program
• NOAA Pacific Tsunami Warning Center
• NOAA Central Tsunami Warning Center
• Smithsonian Global Volcanism Program
• NOAA Space Weather Prediction Center
• National Hurricane Center (NHC)
• National Interagency Fire Center (NIFC)
• NWS Alerts API
• OpenAQ (Global Air Quality)
• DHS National Terrorism Advisory System (NTAS)
• U.S. State Department Travel Advisories
• CISA Cybersecurity Advisories

OpenClaw Integration

Add to your OpenClaw configuration:

{
  "mcpServers": {
    "wems": {
      "command": "python3",
      "args": ["/path/to/wems-mcp-server/wems_mcp_server.py"],
      "env": {
        "WEMS_CONFIG": "/path/to/config.yaml"
      }
    }
  }
}

Use Cases

• 🏢 Enterprise Risk Management: Automated threat assessment for global operations
• 📺 News Organizations: Real-time natural disaster reporting and alerts
• 🔬 Research Institutions: Data collection for scientific analysis
• 🏠 Personal Safety: Location-specific hazard monitoring for families
• 🤖 AI Emergency Response: Integration with disaster response chatbots
• 📱 Alert Systems: Custom notification workflows for critical events

Advanced Configuration

# config.yaml - Full customization example
alerts:
  earthquake:
    min_magnitude: 6.0
    regions: ["US", "Caribbean", "Pacific"]
    webhook: "https://your-endpoint.com/earthquake"
    
  solar:
    min_kp_index: 7  # G3+ geomagnetic storm
    webhook: "https://your-endpoint.com/solar"
    
  volcano:
    alert_levels: ["WARNING", "WATCH"] 
    regions: ["Cascade Range", "Ring of Fire"]
    webhook: "https://your-endpoint.com/volcano"
    
  tsunami:
    enabled: true
    regions: ["pacific", "atlantic", "indian"]
    webhook: "https://your-endpoint.com/tsunami"

Monitoring Dashboard

Pair with monitoring tools for comprehensive coverage:

# Example: Send earthquake data to monitoring system
curl -X POST https://your-monitoring.com/api/events \
  -H "Content-Type: application/json" \
  -d "$(python -c 'import wems; print(wems.get_recent_earthquakes())')"

Billing & Monetization (Current)

WEMS now includes Stripe metering scaffolding and affordable default pricing.

Current pricing defaults

• Free tier: 5,000 calls per rolling 30 days
• 0–100,000 calls: $0.0010/call
• 100,001–500,000 calls: $0.0008/call
• 500,001+ calls: $0.0006/call

Accessory call weights (default)

• Most tools: 1 unit
• check_space_weather_alerts: 2 units
• fuse_multi_source_incidents: 3 units

Billing config

See: config/wems_stripe_billing.json

Key fields:

• event_name
• api_key_to_customer
• billing_units.default
• billing_units.by_tool
• pricing.free_calls_per_rolling_30d
• pricing.tiers[]

Stripe key source

1. STRIPE_API_KEY or STRIPE_SECRET_KEY (direct env)

WEMS uses best-effort lookup and never blocks alerting if billing key resolution fails.

Built with ❤️ for the AI community by Helios 🌞

Part of the expanding OpenClaw ecosystem