Provides real-time hazard data from authoritative sources with configurable alerts and webhooks for AI agents.
Configuration
View docs{
"mcpServers": {
"heliosarchitect-wems-mcp-server": {
"command": "python3",
"args": [
"/path/to/wems-mcp-server/wems_mcp_server.py"
],
"env": {
"WEMS_CONFIG": "/path/to/config.yaml"
}
}
}
}You can deploy this MCP server to connect AI assistants with authoritative natural hazard data sources. It provides real-time monitoring, configurable alerts, and webhook actions across multiple hazard types, enabling you to build proactive risk-aware AI workflows without heavy manual data gathering.
You run the MCP server locally or in your environment and connect your MCP client to it. The server gathers data from trusted sources, evaluates alert thresholds you configure, and triggers webhooks when events meet your criteria. You can tailor monitoring to specific regions or operate globally, and you can add multiple hazard modules such as earthquakes, tsunamis, volcanoes, and space weather.
Prerequisites: Python is required. Ensure you have Python 3.8 or newer installed on your system. You also need network access to reach data sources.
pip install wems-mcp-serverIf you prefer to install from source, clone the repository, install dependencies, and install the package.
git clone https://github.com/heliosarchitect/wems-mcp-server.git
cd wems-mcp-server
pip install -r requirements.txtStart the server and connect it to your MCP client so AI assistants can query and receive alerts. The server can be run as a standalone MCP server and will use the configured alerts and data sources.
# Run the MCP server so it can connect to AI assistants
python -m wems_mcp_serverConfigure alert thresholds, regions, and webhooks to control what you are notified about. The following example shows common sections you may include in your configuration file.
alerts:
earthquake:
min_magnitude: 6.0
regions: ["US", "Caribbean", "Pacific"]
webhook: "https://your-endpoint.com/earthquake"
solar:
min_kp_index: 7
webhook: "https://your-endpoint.com/solar"
tsunami:
enabled: true
regions: ["pacific", "atlantic", "indian"]
webhook: "https://your-endpoint.com/tsunami"After starting, you can query the server from your MCP client to check current alerts, test individual hazard checks, and verify that webhooks are firing as configured. Use the client to request data from specific hazard modules and regions, then integrate the responses into your AI agent prompts or automation pipelines.
The server aggregates data from authoritative sources such as USGS for earthquakes, NOAA centers for tsunami and space weather, Smithsonian for volcanoes, NHC for hurricanes, OpenAQ for air quality, and other national and international authorities. Coverage is configurable by hazard type and region.
Query recent earthquake activity and thresholds across regions to identify significant events.
Monitor space weather indicators such as K-index, solar flares, and coronal mass ejections.
Track volcanic activity alerts from authoritative sources.
Monitor tsunami warnings and related impact assessments.
Track tropical cyclones, forecasts, and potential impacts.
Deliver fire weather alerts and active perimeters.
Monitor severe weather events like tornadoes and flash floods.
Provide flood warnings and river gauge data.
AQI data and pollutant measurements from OpenAQ.
Monitor terrorism, travel risk, and cyber threat advisories.
Active space weather alerts and warnings from NOAA SWPC.
US drought conditions with D0-D4 levels (Premium).
Update alert thresholds and webhook endpoints.