Skip to content
Docs navigation

Agents as developers

Examples

The server's prompts show how its tools compose. Hosts with a prompt picker (Claude Desktop, MCP Inspector) list each one on a connection that serves every tool it calls.

Severe weather briefing

severe_weather_briefing

Full situational-awareness briefing for a location: SPC outlook + narrative, active alerts, hazardous weather outlook, and severe-weather probabilities, synthesized into a meteorologist-style briefing.

Arguments

  • location (required): City/state, ZIP, or "lat,lon" for the briefing location

Served on: /x/core,catalog,guidance or any connection that includes those toolsets

Compare locations for an event

compare_locations_for_event

Side-by-side weather comparison across candidate locations for a planned outdoor event, with a recommendation.

Arguments

  • locations (required): Comma-separated candidate locations (2-5), e.g. "Austin TX, Denver CO"
  • event (required): What is being planned, e.g. "Saturday afternoon wedding"

Served on: /x/core,catalog,planning or any connection that includes those toolsets

Travel weather outlook

travel_weather_outlook

Weather outlook for a trip: conditions at origin, destination, and travel-day impacts (precipitation, wind, aviation conditions when flying).

Arguments

  • origin (required): Departure location
  • destination (required): Arrival location
  • travelDate (required): Travel date (YYYY-MM-DD or natural language)

Served on: / (the root) and every wider connection

Single-call examples

Input examples the tools publish with their schemas. Each tool's full parameters are in the tool reference.

search_locationscore
{
  "query": "Springfield"
}
{
  "query": "bost",
  "fuzzy": true,
  "limit": 3
}
{
  "query": "50219"
}
get_observed_precipitationcore
{
  "location": "Chicago, IL",
  "windows": [
    "24h"
  ]
}
{
  "lat": 41.89,
  "lon": -87.65,
  "windows": [
    "1h",
    "6h"
  ],
  "at": "2026-09-30T12:00:00Z"
}
search_catalogcore
{
  "query": "how much rain fell in the last day",
  "lat": 39.74,
  "lon": -104.99,
  "limit": 3
}
find_best_windowplanning
{
  "location": "Boulder, CO",
  "criteria": {
    "min_temperature": 55,
    "max_wind_speed": 15,
    "max_precipitation_probability": 20
  },
  "hours": 72,
  "activity_duration_hours": 3
}
{
  "location": "Girdwood, AK",
  "criteria": {
    "max_precipitation_probability": 30
  },
  "daylight_only": true
}
get_forecast_skillverification
{
  "location": "Denver",
  "variable": "temperature_2m",
  "lead_hours": 72
}
{
  "lat": 39.7392,
  "lon": -104.9903
}
{
  "location": "Dallas, TX",
  "variable": "precipitation_probability",
  "lead_hours": 24
}
{
  "location": "Chicago",
  "variable": "temperature_2m",
  "regime": "SEA:DJF"
}
{
  "location": "Houston",
  "variable": "precipitation_gt_0p254mm",
  "lead_hours": 24
}
get_forecast_skill_mapverification
{
  "bbox": {
    "west": -105.2,
    "south": 39.5,
    "east": -104.7,
    "north": 40
  },
  "model": "nbm",
  "variable": "temperature_2m",
  "lead_hours": 24
}
get_map_snapshotvisual
{
  "location": "Tulsa, OK",
  "product": "nexrad_l3:TLX:N0B",
  "zoom": 8
}
{
  "lat": 39.74,
  "lon": -104.99,
  "product": "goes_truecolor_conus",
  "zoom": 5
}
{
  "location": "Chicago, IL",
  "product": "hrrr_precip_hybrid_derived_conus",
  "zoom": 7,
  "time": 1783554900
}

Tips for Better Results

  • Be specific about location. Include city and state to avoid ambiguity — “Portland, OR” vs. “Portland, ME”.
  • Specify time windows. “This weekend”, “next 6 hours”, or “Tuesday through Thursday” helps the AI choose appropriate parameters.